System and method for communication between, and control of, a plurality of vehicles having autonomous driving functions or trajectory predictive driving functions, and / or one or more objects capable of being connected to a remote server

Abstract

A system is described for communication between, and control of, a plurality of vehicles equipped with autonomous driving functions or trajectory predictive driving functions, and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personal electronic device or drones. The electronic controller (3) of each vehicle is configured to continuously transmit data related to the geographical position and to the operative conditions of the vehicle, the information indicating the cognitive state, the intentions, and the emotions of the driver to a remote server (C) that builds a metaverse consisting of a three-dimensional virtual simulation of the scenario in which the autonomous driving vehicles and / or the connected objects move, including a three-dimensional representation of the road network, configured so as to enable distinguish between road routes that are adjacent to each other and / or that intersect with each other but are located at different heights. In said metaverse, a digital twin (HV', RV') of each vehicle (HV, RV) and of each connected object is created, and for each of such vehicles, a respective most probable path is calculated and associated therewith, and consequently, a three-dimensional volume of relevance (S') used to filter out elements and / or events that are of relevance for the vehicle as they are contained within the three-dimensional volume of relevance of the vehicle. In case a risk of intersection between volumes of relevance (S') of different vehicles is detected in the metaverse, a feedback message is sent to one or more vehicles (HV, RV).

Description

[0001] "System and method for communication between, and control of, a plurality of vehicles having autonomous driving functions or trajectory predictive driving functions, and / or one or more objects capable of being connected to a remote server"

[0002] ****

[0003] TEXT OF THE DESCRIPTION

[0004] Technical field

[0005] The embodiments of the present description relate to systems and methods for controlling autonomous driving vehicles.

[0006] In particular, various embodiments of the present description concern systems and methods for communication between, and control of, a plurality of vehicles having autonomous driving functions or trajectory predictive driving functions, and / or one or more objects capable of being connected to a remote server (such as pedestrians equipped with a personal electronic device, drones, or similar).

[0007] Technical background

[0008] In the field of autonomous driving road vehicles, significant developments have been underway for several years. A common criterion for describing the autonomy level of an autonomous driving vehicle is based on a SAE ("Society of Automotive Engineering") classification from 0 to 5. The most modem cars equipped with devices such as adaptive cruise control and / or automatic lane centering systems are classified at level L1 . Level L2 autonomous vehicles may include, for example, more than a dozen sensor devices. In general, vehicles between levels L0 and L3 exhibit increasing features of comfort and safety but still require the driver to remain alert and ready to intervene if necessary. A major leap occurs at level L4, with vehicles that can be used in limited areas without any driver intervention. At level L5, a car may be equipped with more than thirty additional sensors to achieve complete driving control substantially without human intervention. Level L5 represents full autonomy, wherein the vehicle is capable of operating without any driver intervention in any environment (urban, suburban, rural, highway, and even off-road) at any reasonable speed and under any environmental condition.

[0009] The present solution particularly relates to autonomous driving vehicles, at least of level L1 , with autonomous driving functions or trajectory planning functions, and to a communication and control system for such vehicles and / or objects capable of being connected to a remote server (such as pedestrians equipped with a personal electronic device, drones, etc.). The autonomous driving vehicles are of the type comprising: a plurality of sensors for perceiving the external environment of the vehicle, for example, sensors included in an ADAS ('Advanced Driver-Assistance System") capable of reconstructing the environment around the vehicle through object detection, a plurality of sensors for detecting the operating conditions of the vehicle, at least one position sensor for determining the current geographical position of the vehicle, a vehicle driving system, an electronic controller configured to receive output signals from said sensors and to issue commands to said vehicle driving system, a communication module connected to said electronic controller and configured to communicate with a remote server and with the communication modules of other autonomous driving vehicles traveling near the vehicle.

[0010] Said communication module is generally capable of transmitting V2X ("Vehicle-to-Everything") messages and typically may use a C-ITS system ("Cooperative Intelligent Traffic System"), as per ETSI standards, configured to transmit CAM ("Cooperative Awareness Message") messages, for example, according to the ETSI EN 302 637-2 standard, and / or CPM ("Collective Perception Messages," according to the ETSI TR 103 562 standard).

[0011] Figure 1 of the attached drawings shows a schematic of an autonomous driving vehicle of the type described above.

[0012] With reference to Figure 1 , the number 1 indicates as a whole an autonomous driving vehicle of any known type, including a level L1 autonomous driving system 2, configured to control, in a known manner, all operational components of the vehicle (engine, transmission, steering, accelerator, brakes, etc.). The level L1 autonomous driving system 2 is controlled by an electronic controller 3 based on signals received from a suite of sensors 4 of any known type (video cameras, radar sensors, lidar sensors, etc.), capable of perceiving the environment surrounding the vehicle, and from a suite of sensors 5 configured to detect a plurality of parameters indicative of the operating conditions of the vehicle and its components, such as distance traveled, speed, longitudinal, transverse, and vertical accelerations, directional changes, etc. The vehicle is also equipped with a position sensor 6 of any known type capable of detecting the geographical position of the vehicle. Associated with the controller 3 are one or more memories M containing geolocalized data (in the figure, for simplicity, the case of a single memory M is illustrated).

[0013] The electronic controller 3 is also connected to a communication module T of any known type, capable of transmitting V2X ("Vehicle-to- Everything") messages to the environment outside of the vehicle, particularly to a remote server C and / or to other vehicles. The transmitted V2X messages are specifically broadcast messages of the CAM ("Cooperative Awareness Message") type and / or messages of the CPM ("Collective Perception Messages" according to the ETSI TR 103 562 standard). The communication module T is also configured to exchange both CAM-type messages and CPM-type messages, as well as DENM ("Decentralized Environmental Notification Message," according to the ETSI TS 102 637-2 standard) messages.

[0014] With reference to Figure 2, which illustrates a communication network to which the vehicles are connected, the electronic controller 3 of each vehicle is configured to send both messages to a cloud service server C and messages to a MEC ("Multi-access Edge Computing") server, via base stations BS (Base Station), as well as V2V ("Vehicle-to-Vehicle") messages to other vehicles and V2N ("Vehicle-to-Network") messages to BS nodes.

[0015] According to a conventional technique, each MEC server can enable efficient network operation, by being connected to a base station BS near a road. Vehicles traveling along the road can connect both to an adjacent MEC server and to the server C. Recently, a "GeoNetworking" technique (according to the ETSI EN 302 636-4-1 standard) has been developed for communication between vehicles (V2V) and between vehicles and infrastructures (V2I) within an ITS ("Intelligent Transport Systems"). The GeoNetworking system uses the geographical positions of the vehicles to determine, for each vehicle, a two- dimensional relevance area around each vehicle, which is used to filter elements and / or events that are relevant to the vehicle, as they are contained within said relevance area. A limitation of this system lies in the fact that it is primarily based on two-dimensional data, which, for example, prevents distinguishing between road paths that are adjacent to each other or that intersect with each other but are at different levels, so that vehicles traveling on a road path at a different level from the reference vehicle's road path erroneously appear, in a two-dimensional representation, as vehicles that may interfere along the reference vehicle's path.

[0016] There is therefore a need for further improvements in this field.

[0017] Object and summary

[0018] An object of one or more embodiments is to implement a system and method for communication between and control of a plurality of vehicles and / or connected objects, equipped with autonomous driving functions or trajectory predictive driving functions, which allow for safer and more efficient identification and reporting of all elements and / or events that are relevant to each of the controlled vehicles.

[0019] In particular, an object of the invention is to more reliably and efficiently filter, for each vehicle, the elements (people and / or vehicles and / or infrastructures) and / or events that may have an actual impact on the vehicle.

[0020] According to one or more embodiments, the aforementioned object may be achieved by a system for communication between, and control of, a plurality of vehicles equipped with autonomous driving functions or trajectory predictive driving functions, and / or one or more objects capable of being connected to a remote server, having the features set forth in the claims that follow, which constitute an integral part of the technical teaching provided herein.

[0021] Therefore, the present detailed description relates to a system for communication between and control of a plurality of vehicles having autonomous driving functions or path predictive driving functions and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personal electronic device or drones, wherein each autonomous driving vehicle includes: a plurality of sensors for detecting the environment outside of the vehicle, a plurality of sensors for detecting the operating conditions of the vehicle, a position sensor for determining the current geographical position of the vehicle, one or more sensors of vital parameters of the driver of the vehicle, including EEG (electroencephalogram) sensors configured to monitor the brain activity of the driver, a vehicle driving system, an electronic controller configured to receive output signals from said sensors and to send commands to said vehicle driving system, and a communication module, connected to said electronic controller and configured to communicate with said remote server and with the communication modules of other autonomous driving vehicles traveling in the vicinity of the vehicle and / or with personal electronic devices of vulnerable subjects present in the vicinity of the vehicle, wherein the electronic controller of each autonomous driving vehicle is configured to process the output signals from the EEG sensors to determine the following information: the cognitive state of the driver, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the driver, and the emotions of the driver, including in particular a state of sudden fear due to a perception of danger, wherein the electronic controller of each autonomous driving vehicle is configured to continuously transmit data related to the geographical position of the vehicle, data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and the information indicating the cognitive state, the intentions, and the emotions of the driver to said remote server, wherein the remote server is configured to build a metaverse consisting of a three-dimensional virtual simulation of a scenario comprising the autonomous driving vehicles and said objects which are connected to said remote server, including a three-dimensional representation of a road network, configured so as to enable distinguishing between road routes that are adjacent to each other and / or that intersect with each other but are located at different heights, and further including a representation of infrastructures and / or objects that are located along and adjacent to the road network, said remote server being further configured to perform, continuously and cyclically, the following operations: receive from the electronic controller of each autonomous driving vehicle said data related to the geographical position of the vehicle, said data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and said information indicating the cognitive state, the intentions, and the emotions of the driver, determine, based on the data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, a most probable path of each autonomous driving vehicle, correct said most probable path of each autonomous driving vehicle based on the information indicating the cognitive state, the intentions, and the emotions of the vehicle driver, obtaining a corrected most probable path, create, in said metaverse, a digital twin of each vehicle and / or connected object, associating a respective corrected most probable path with each digital twin of a vehicle, create, for each digital twin of a vehicle in said metaverse, based on the corrected most probable path associated with each digital twin of a vehicle, a three-dimensional volume of relevance which is used to filter out elements and / or events that are of relevance for each vehicle, as they are contained within said three-dimensional volume of relevance, said three-dimensional volume of relevance that is associated with each digital twin of a vehicle being a volume that extends along said corrected most probable path calculated for the vehicle and that has a cross-section dimensioned to include therewithin it the cross-section of the vehicle, detect, in said metaverse, any risk of intersection between the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volumes of relevance of the digital twins of other vehicles or with representations of infrastructures or objects located along or adjacent to the road network, and when a risk of intersection is detected, send a feedback message from said remote server to one or more vehicles or objects to signal that a risk of intersection has been detected, when a risk of intersection is signaled, define, in the electronic controllers of one or more of said vehicles or objects, actions to reduce or eliminate said risk of intersection detected.

[0022] In a preferred embodiment, said one or more objects capable of being connected to the remote server comprise at least one vulnerable subject located adjacent to or along the road, outside the autonomous driving vehicles, preferably a pedestrian, in particular a child or elderly person, a cyclist or motorcyclist, or disabled subjects in wheelchairs, said at least one vulnerable subject being equipped with a personal electronic device, said system further comprising: one or more sensors of vital parameters of said at least one vulnerable subject, including EEG sensors configured to monitor the brain activity of said at least one vulnerable subject, and said personal electronic device, preferably a smartphone or tablet in the equipment of the at least one vulnerable subject, configured to process the output signals from the EEG sensors of said at least one vulnerable subject to determine the following information: the cognitive state of said at least one vulnerable subject, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of said at least one vulnerable subject, and the emotions of said at least one vulnerable subject, including in particular a state of sudden fear due to a perception of danger, wherein said personal electronic device of said at least one vulnerable subject is configured to continuously transmit the information indicating the cognitive state, the intentions, and the emotions of said at least one vulnerable subject to said remote server directly and / or via electronic controllers of autonomous driving vehicles or of infrastructures located within a determined distance from said at least one vulnerable subject, wherein said operations performed continuously and cyclically by the remote server further comprise: receive from the personal electronic device of said at least one vulnerable subject said information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject, determine, based on said information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject, a most probable path of said at least one vulnerable subject, create, in said metaverse, a digital twin of said at least one vulnerable subject by associating it with said most probable path of said at least one vulnerable subject, create for said digital twin of said at least one vulnerable subject in said metaverse, based on said most probable path of said at least one vulnerable subject, a three-dimensional volume of relevance that is used to filter out elements and / or events that are of relevance for said at least one vulnerable subject, as they are contained within said three- dimensional volume of relevance, said three-dimensional volume of relevance that is associated with the digital twin of said at least one vulnerable subject being a volume that extends along said determined most probable path of said at least one vulnerable subject and that has a crosssection dimensioned to include therewithin it the cross-section of said at least one vulnerable subject, detect, in said metaverse, any risk of intersection between the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volume of relevance of said at least one vulnerable subject, and when a risk of intersection is detected, send a feedback message from said remote server to one or more vehicles and / or to said at least one vulnerable subject to signal that a risk of intersection has been detected.

[0023] Still more preferably, said personal electronic device of said at least one vulnerable subject comprises a position sensor configured to determine the current geographical position of said at least one vulnerable subject, said personal electronic device of said at least one vulnerable subject is configured to continuously transmit data related to the geographical position of said at least one vulnerable subject to said remote server directly and / or via electronic controllers of autonomous driving vehicles or of infrastructures located within a determined distance from said at least one vulnerable subject, and said operation of determining, based on said information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject, the most probable path of said at least one vulnerable subject is further performed based on said data related to the geographical position of said at least one vulnerable subject.

[0024] According to a further preferred characteristic, the personal electronic device of said at least one vulnerable subject is configured to continue to transmit dynamically to said remote server, directly and / or via electronic controllers of autonomous driving vehicles or of infrastructures located within a determined distance from said at least one vulnerable subject, the information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject, for the repetition by the remote server of said operations that the remote server performs continuously and cyclically.

[0025] Always preferably, the electronic controller of each autonomous driving vehicle is configured to continue to transmit dynamically to said remote server the data related to the geographical position of the vehicle, to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and the information indicating the cognitive state, intentions, and emotions of the driver for the repetition by the remote server of said operations that the remote server performs continuously and cyclically.

[0026] One or more embodiments relate to a corresponding method.

[0027] As evident from the foregoing, a fundamental advantage of the present invention lies in the fact that the processing operations necessary to filter the relevant elements and events at the network level, that is, considering the "Network Layer" of the ISO / OSI model, for each vehicle or object in the system are performed in the remote server, based on a "metaverse" consisting of a three-dimensional virtual simulation of a real scenario including the vehicles and objects. The term "metaverse" is precisely used, both in the present description and in the claims that follow, to indicate said three-dimensional virtual simulation of the real scenario.

[0028] Brief description of the figures

[0029] One or more embodiments will now be described, by way of example only, with reference to the attached figures, wherein:

[0030] Figure 1 is a schematic view of an autonomous driving vehicle;

[0031] Figure 2 illustrates the architecture of a communication network among multiple autonomous driving vehicles;

[0032] Figure 3 shows the driver of an autonomous driving vehicle forming part of the system according to the solution, with EEG ("Electroencephalogram") sensors for detecting the brain activity of the driver which in this specific example are associated with a headgear (alternatively, EEG sensors associated with earphones or similar may be used);

[0033] Figure 4 shows a pedestrian located adjacent to a road traveled by one or more autonomous vehicles, equipped with EEG sensors for detecting brain activity, for example associated with a headset or earphones worn by the pedestrian, and further provided with a portable electronic device, for example of the smartphone or tablet type, in communication with a remote server and / or with electronic controllers of autonomous vehicles located nearby;

[0034] Figure 5 shows an example of a situation where EEG sensors are used to detect the brain activity of vulnerable users, for example, pedestrians about to cross a road while an autonomous vehicle is approaching;

[0035] Figures 6 to 9 are schematic plan and elevation views presenting different scenarios where multiple autonomous driving vehicles move, showing the advantages of the present invention over known methods; Figure 10 is a schematic plan view where an autonomous driving vehicle and a vulnerable subject move, showing the advantages of the present invention over known methods;

[0036] Figure 11 is a schematic illustrating the operation of the system according to the solution; and

[0037] Figures 12 and 13 are flow diagrams illustrating an exemplary embodiment of the method according to the solution.

[0038] The figures are reproduced to clearly illustrate the relevant aspects of the embodiments and are not necessarily drawn to scale.

[0039] Detailed description of example embodiments

[0040] In the following description, one or more specific details are illustrated for the purpose of providing an in-depth understanding of example embodiments of this description. The embodiments may be obtained without one or more of the specific details or with other methods, components, materials, etc. In other cases, known operations, materials, or structures are not illustrated or described in detail so that certain aspects of the embodiments will not be obscured.

[0041] A reference to "an embodiment" within the context of the present description is intended to indicate that a particular configuration, structure, or feature described with reference to the embodiment is included in at least one embodiment. Therefore, phrases such as "in one embodiment" or similar that may appear in one or more points of the present description do not necessarily refer to the same embodiment.

[0042] Moreover, particular configurations, structures, or features may be combined in any suitable manner in one or more embodiments.

[0043] The references used herein are provided merely for convenience and thus do not define the scope of protection or the scope of the embodiments.

[0044] In all the figures attached herein and throughout the following detailed description, unless the context indicates otherwise, similar parts or elements are indicated with similar references / numbers and a corresponding description will not be repeated for brevity.

[0045] Figures 1 and 2 have already been described above. Essentially, Figure 1 shows a schematic of an autonomous driving vehicle, while Figure 2 shows a communication system among multiple autonomous driving vehicles of a type to which the present invention may be applied.

[0046] Solutions according to the present description refer to vehicles according to Figure 1 further comprising at least one device for detecting the brain activity of the driver.

[0047] Therefore, with reference to Figure 3, a device 8 for detecting the brain activity of the seat occupant is associated with the driver's seat 7 in the cabin of the autonomous driving vehicle.

[0048] In the solution illustrated herein by way of non-limiting example, the device 8 includes a plurality of EEG sensors 9, carried by a temporal band 10 and two parietal bands 11 forming part of a headset structure, covered by a headgear 12 (shown with a dashed line). The sensors 9 are connected to a bundle of cables 13 that is received inside the structure of the seat 7 (where an elastic retraction reel 14 is provided) and that is connected to the electronic controller 3. In the specific example illustrated herein, a hook 15 is provided on one side of the seat 7 to receive the device 8 when it is not in use. This solution is merely an example. Any alternative solution capable of associating EEG sensors to the driver and vehicle occupants may be used. For example, wearable devices such as ear sensors that can be positioned in the users' ears and capable of detecting the brain activity of the users have already been proposed. Any other known solution may be used for this purpose.

[0049] Note that, in solutions according to the present description, additional devices 8 for detecting brain activity may be present on the other seats in the cabin of the autonomous driving vehicle.

[0050] In general, the solution relates to a system that controls a plurality of autonomous driving vehicles, for example of the type illustrated in Figure 3, which are in communication with each other, for example in a network of the type illustrated in Figure 2.

[0051] Moreover, solutions according to the present description may relate to a system capable of controlling such plurality of autonomous driving vehicles (of the type illustrated in Figure 3) based on: both, as already described, communications exchanged among them, for example, via the network of the type illustrated in Figure 2 and via a remote server C, and communications exchanged between such plurality of autonomous driving vehicles and vulnerable subjects (VRU), for example, again via the network of the type illustrated in Figure 2 and via the remote server C, which are located adjacent to or along a road, outside the autonomous vehicles, such as pedestrians, in particular children and elderly persons, or cyclists and motorcyclists, or disabled persons in wheelchairs.

[0052] In this case, these vulnerable subjects are also equipped with respective brain activity monitoring devices 8, thereby obtaining a plurality of EEG sensors for brain activity associated with a plurality of VRUs located adjacent to or along the road, outside the autonomous vehicles.

[0053] By way of example only, Figure 4 shows a VRU pedestrian located on a sidewalk M adjacent to a road and near a traffic light device F. The VRU subject wears a headset 16 bearing a plurality of EEG sensors 9 for monitoring the brain activity of the VRU subject. Naturally, as already indicated above, the EEG sensors could be associated with the subject in different ways, for example they could be associated with earphones inserted into the subject's ears.

[0054] Still referring to the example illustrated in Figure 4, the EEG sensors 9 are connected, either wired or wirelessly (e.g., via Bluetooth), to a portable electronic device 17 carried by the VRU subject, such as a smartphone or tablet device, which is in communication with the remote server C.

[0055] Figure 5 shows another situation where two VRU subjects, specifically a child and an elderly person, are about to cross a road R at a pedestrian crossing CR while an autonomous vehicle 1 is approaching. Both the electronic devices carried by the VRU subjects and the electronic controller of the autonomous vehicle 1 are in communication with the remote server C. Also in communication with the remote server C is an electronic controller of an infrastructure F, specifically a traffic light device.

[0056] In the system according to the solution, the electronic controller 3 of each autonomous vehicle is configured to continuously transmit to the server C, via the vehicle's communication module T, data related to the geographical position of the vehicle and data related to the vehicle's state such as speed and dynamics data, for example, according to the structure of the previously described CAM messages.

[0057] Furthermore, in solutions according to the present description, such electronic controller 3 of each autonomous vehicle is configured to process the output signals from the EEG sensors 9 associated with the driver to determine the following information:

[0058] - the cognitive state of the driver, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness,

[0059] - the intentions of the driver, and

[0060] - the emotions of the driver, including in particular a state of sudden fear due to a perception of danger.

[0061] Additionally, such electronic controller 3 is configured to continuously transmit to the server C, via the vehicle's communication module T, the information determined by processing the output signals from the EEG sensors 9 associated with the driver.

[0062] Note that by transmitting to the server C this information determined by processing the output signals from the EEG sensors 9 associated with the driver, it is possible to reduce road accidents caused by misinterpretations of other drivers' behaviors.

[0063] The structure through which the information determined by processing the output signals from the EEG sensors 9 associated with the driver (or VRUs as described below) is transmitted to the server C can be optimized to increase bandwidth efficiency and reduce latency, achieving low-latency communications.

[0064] For example, this information can be transmitted via a JSON ("JavaScript Object Notation") message, hereinafter referred to as an EEGM (Electroencephalography Message), comprising a first "header" field containing metadata about the message itself and a second "payload" field containing the actual data related to the information to be transmitted, i.e., the EEG data.

[0065] The "header" field may include, for example: a message ID, for example labeled "messageld", containing a unique identifier for the message; a timestamp, for example labeled "timestamp", containing an indicator of the temporal moment when the cognitive state, intention, and / or emotion was detected; a message type, for example labeled "messageType", containing an indicator specifying that the message relates to the cognitive state, intentions, and / or emotions of the driver or a vulnerable subject, i.e., that the message is an EEGM message as described here.

[0066] The "payload" field may include, for example: a driver or vulnerable subject ID, for example labeled "driverld", containing a unique identifier for the driver or vulnerable subject; an action, for example labeled "intent", containing an element specifying the type of action the driver or vulnerable subject wants to perform, such as turning, proceeding straight, or similar; a probability, for example labeled "probabilities", containing a probability for each possible action that can be performed by the driver or vulnerable subject, such as: a probability associated with a right turn, for example labeled "right", a probability associated with a left turn, for example labeled "left", and a probability associated with proceeding straight, for example labeled "straight"; and a confidence, for example labeled "confidence", containing an element with a value between 0 and 1 indicating the model's confidence in the action prediction.

[0067] An exemplary EEGM message may be as follows:

[0068] {

[0069] "header": {

[0070] "messageld": "EEG01",

[0071] "timestamp": "2024-05-14T14:37: 12+02:00",

[0072] "messageType": "Driverintent"

[0073] },

[0074] "payload": {

[0075] "driverld": "DR12345",

[0076] "intent": "TurnSignal",

[0077] "probabilities": {

[0078] "right": 0.60,

[0079] "left": 0.25,

[0080] "straight": 0.15

[0081] },

[0082] "confidence": 0.95

[0083] }

[0084] } Note that in the exemplary message above, the "header" field includes: the message ID "messageld", which assigns to the message a unique identifier with the value "EEG01"; the timestamp "timestamp", which assigns to the message an indicator of the temporal moment when the cognitive state, intention, and / or emotion of the driver or vulnerable subject was detected, i.e., the value "2024-05-14T14:37: 12+02:00". Note that the format of the temporal indication may vary from the one shown here; the message type "messageType", which, via the value "Driverintent", indicates that the message relates to the cognitive state, intentions, and / or emotions of the driver or vulnerable subject, i.e., that the message is an EEGM message.

[0085] The "payload" field instead includes: the driver or vulnerable subject ID "driverld", which assigns the driver or vulnerable subject a unique identifier with the value "DR12345"; the action "intent", which specifies the type of action the driver or vulnerable subject wants to perform, i.e., in the example described here, a turn action indicated via the label "TurnSignal"; the probability "probabilities", containing the probability for each possible action (e.g., in the example considered, turning right, turning left, and proceeding straight) that can be performed by the driver or vulnerable subject, i.e.: a 60% (0.60) probability of turning right, indicated via the label "right", a 25% (0.25) probability of turning left, indicated via the label "left", and a 15% (0.15) probability of proceeding straight, indicated via the label "straight"; and the confidence "confidence", which indicates a confidence value of the model in the action prediction of 0.95.

[0086] The described structure allows efficient encoding of the required information, generating a message suitable for low-bandwidth, low-latency V2X communication. Note that the fields can be adapted based on the specific requirements and constraints of the system in use, so some of the described fields may not be present or additional fields may be included. If the control of the plurality of autonomous driving vehicles is also performed based on communications exchanged between autonomous driving vehicles and vulnerable road users (VRUs) located adjacent to or along a road, outside the autonomous vehicles, the server C is configured to continuously receive information determined by processing the output signals from the EEG sensors 9 associated with the VRUs.

[0087] Indeed, the portable electronic device 17 associated with each VRU can be configured to process the output signals from the EEG sensors 9 of the vulnerable subject to determine the following information: the cognitive state of the VRU, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the VRU, the emotions of the VRU, including in particular a state of sudden fear due to a perception of danger.

[0088] This information can then be transmitted to the server C directly or via an electronic controller of an infrastructure F or an electronic controller 3 of an autonomous driving vehicle.

[0089] Note that by transmitting to the server C such information determined by processing the output signals from the EEG sensors 9 associated with vulnerable subjects, for example pedestrians, it is possible to reduce road accidents caused by misinterpretations of the behaviors of such vulnerable subjects.

[0090] Note that to control the plurality of autonomous driving vehicles based on communications exchanged, for example via the remote server C, between autonomous driving vehicles and vulnerable subjects (VRUs), the portable electronic device 17 provided to each VRU may be equipped with an integrated position sensor. In this case, the electronic device 17 can be configured to communicate the position of the VRU detected by the position sensor to the remote server C and / or to the electronic controller 3 of autonomous vehicles located within a determined distance from the VRU.

[0091] The server C is configured to build a metaverse consisting of a three- dimensional virtual simulation of the scenario in which the autonomous driving vehicles move, including a three-dimensional representation of the road network, configured to distinguish between road paths that are adjacent to each other and / or intersect with each other but are located at different heights, and further including a representation of infrastructures and / or objects located along and adjacent to the road network.

[0092] As already indicated, the term "metaverse" is used, both in the present description and in the claims that follow, to refer to said three- dimensional virtual simulation of the real scenario in which the autonomous driving vehicles move.

[0093] This three-dimensional virtual simulation of the scenario is used within the server C to: determine, based on data related to the vehicle's state such as speed and data related to the vehicle's dynamics received from the electronic controllers 3 of the autonomous driving vehicles, a most probable path for each vehicle; correct, based on the information determined by processing the output signals from the EEG sensors 9 associated with the driver of each vehicle, such most probable path of each vehicle, obtaining a corrected most probable path for each vehicle based on the information related to the brain activity of the respective driver, that is, a most probable path that takes into account the cognitive state, intentions, and emotions of the driver; and estimate, based on such corrected most probable paths as a function of the brain activity of the respective driver and the data related to the geographical position of each vehicle, a three-dimensional relevance volume for each of the vehicles, that is, a simulated three-dimensional path ("3D Path") having a corresponding reference volume.

[0094] If the control of the plurality of autonomous driving vehicles is also performed based on communications exchanged between autonomous driving vehicles and vulnerable subjects (VRUs) located adjacent to or along a road, outside the autonomous vehicles, such three-dimensional virtual simulation of the scenario is used within the server C to estimate, based on data received from the electronic controllers 3 of the autonomous driving vehicles and based on the information determined by processing the output signals from the EEG sensors 9 associated with each of the VRUs, both a three-dimensional volume of relevance for each of the vehicles, that is, a simulated three-dimensional path ("3D Path") having a corresponding volume of reference related to the vehicles, and a three-dimensional volume of relevance for each of the vulnerable subjects (VRUs), that is, a simulated three-dimensional path ("3D Path") having a corresponding volume of reference related to the vulnerable subjects.

[0095] Note that such three-dimensional relevance volume estimated for each of the vulnerable subjects (VRUs) can be obtained based on a most probable path of the respective vulnerable subject determined as a function of the brain activity of that respective vulnerable subject, that is, based on their cognitive state, intentions, and emotions, and, preferably, their current position.

[0096] Furthermore, such server C can be configured to create and process, based on data received from objects present in the real scenario, for example from intelligent road structures, three-dimensional volumes of relevance related to such objects, for example, to which a digital twin and a corresponding volume of reference are associated.

[0097] In this way, the three-dimensional virtual simulation of the scenario can also include a three-dimensional simulation of the objects present in the simulated environment and a three-dimensional simulation of the movement of each of the vehicles and, if considered, of each of the vulnerable subjects.

[0098] Therefore, the server C can be configured to detect intersections between the three-dimensional volumes of relevance, in particular, intersections between a three-dimensional volume of relevance associated with a digital twin of a vehicle and other three-dimensional volumes of relevance present in the simulation associated with a digital twin of a vehicle, a digital twin of a vulnerable subject, or a digital twin of an object present in the environment, and, in response to detecting an intersection, the server C can be configured to send warning messages indicating a dangerous situation to the vehicles, vulnerable subjects, and / or objects involved in the intersection.

[0099] Thus, solutions according to the present description relate to a system for communication between and control of a plurality of vehicles equipped with autonomous driving functions or trajectory (path) predictive driving functions and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personal electronic device or drones.

[0100] In solutions according to the present description, each autonomous driving vehicle includes: a plurality of sensors 4 for detecting the environment outside of the vehicle, a plurality of sensors 5 for detecting the operating conditions of the vehicle, a position sensor 6 for determining the current geographical position of the vehicle, one or more sensors 9 of vital parameters of the vehicle driver, including EEG (electroencephalogram) sensors 9 configured to monitor the brain activity of the driver, a vehicle driving system 2, an electronic controller 3 configured to receive output signals from such sensors and to send commands to said vehicle driving system 2, and a communication module T, connected to said electronic controller 3 and configured to communicate with said remote server C and with the communication modules T of other autonomous driving vehicles traveling in the vicinity of the vehicle and / or with personal electronic devices of vulnerable subjects present in the vicinity of the vehicle.

[0101] The electronic controller 3 of each autonomous driving vehicle is configured to process the output signals from the EEG sensors 9 to determine the following information: the cognitive state of the driver, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the driver, and the emotions of the driver, including in particular a state of sudden fear due to a perception of danger.

[0102] The electronic controller 3 of each autonomous driving vehicle is further configured to continuously transmit data related to the geographical position of the vehicle, data related to the operative conditions of the vehicle including data related to a state and dynamics of the vehicle, and the information indicating the cognitive state, intentions, and emotions of the driver to said remote server C.

[0103] The remote server C is configured to build a metaverse consisting of a three-dimensional virtual simulation of a scenario comprising the autonomous driving vehicles and said objects connected to said remote server, including a three-dimensional representation of a road network, configured so as to enable distinguishing between road paths that are adjacent to each other and / or that intersect with each other but are located at different heights, and further including a representation of infrastructures and / or objects located along and adjacent to the road network.

[0104] Said remote server C is further configured to perform, continuously and cyclically, the following operations: receive from the electronic controller 3 of each autonomous driving vehicle said data related to the geographical position of the vehicle, said data related to the operative conditions of the vehicle including data related to a state and dynamics of the vehicle, and said information indicating the cognitive state, intentions, and emotions of the driver, determine, based on the data related to the operative conditions of the vehicle including data related to a state and dynamics of the vehicle, a most probable path ("Most-Probable-Path" - MPP) of each autonomous driving vehicle, modify, for example, via a correction operation, said most probable path MPP of each autonomous driving vehicle based on the information indicating the cognitive state, intentions, and emotions of the vehicle driver, obtaining a corrected most probable path, i.e., a most probable path that takes into account the brain activity of the respective driver, create, in said metaverse, a digital twin of each vehicle and / or connected object, associating a respective corrected most probable path with each digital twin of a vehicle, i.e. , one that accounts for the brain activity of the respective driver, create, for each digital twin of a vehicle in said metaverse, based on the corrected most probable path associated with each digital twin of a vehicle, a three-dimensional volume of relevance S' that is used to filter out elements and / or events that are of relevance for each vehicle, as they are contained within said three-dimensional volume of relevance, said three-dimensional volume of relevance S' that is associated with each digital twin of a vehicle being a volume that extends along said corrected most probable path calculated for the vehicle and that has a cross-section dimensioned to include therewithin the cross-section of the vehicle, detect, in said metaverse, any risk of intersection between the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volumes of relevance of the digital twins of other vehicles or with the representations of infrastructures or objects located along or adjacent to the road network, and

[0105] - when a risk of intersection is detected, send a feedback message from said remote server C to one or more vehicles or objects to signal that a risk of intersection has been detected, when a risk of intersection is signaled, define, in the electronic controllers of one or more of said vehicles or objects, actions to reduce or eliminate said risk of intersection detected.

[0106] Furthermore, solutions according to the present description may relate to (if the control of the plurality of autonomous driving vehicles is performed considering vulnerable subjects VRU located adjacent to or along a road, outside of the autonomous vehicles) a system wherein said one or more objects capable of being connected to the remote server include at least one vulnerable subject VRU located adjacent to or along the road, outside the autonomous driving vehicles, for example, a pedestrian such as a child or elderly person, a cyclist or motorcyclist, or a disabled person in a wheelchair.

[0107] Said at least one vulnerable subject VRU is equipped with a personal electronic device 17, and said system further comprises: one or more sensors 9 of vital parameters of said at least one vulnerable subject VRU, including EEG sensors 9 configured to monitor the brain activity of said at least one vulnerable subject VRU, and said personal electronic device 17, preferably a smartphone or tablet provided to said at least one vulnerable subject VRU, configured to process the output signals from the EEG sensors 9 of said at least one vulnerable subject VRU to determine the following information: the cognitive state of said at least one vulnerable subject VRU, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of said at least one vulnerable subject VRU, and the emotions of said at least one vulnerable subject VRU, including in particular a state of sudden fear due to a perception of danger. The personal electronic device 17 of said at least one vulnerable subject VRU may be configured to continuously transmit the information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU to said remote server C, for example, directly, via electronic controllers 3 of autonomous driving vehicles located within a determined distance from said at least one vulnerable subject VRU, and / or via electronic controllers of intelligent infrastructures F located within said determined distance from said at least one vulnerable subject VRU.

[0108] In embodiments considering vulnerable subjects, the operations performed continuously and cyclically by the remote server C may further include: receive from the personal electronic device 17 of said at least one vulnerable subject VRU said information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU, determine, based on said information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU, a most probable path of said at least one vulnerable subject VRU, create, in said metaverse, a digital twin of said at least one vulnerable subject VRU by associating it with said most probable path of said at least one vulnerable subject VRU, create for said digital twin of said at least one vulnerable subject VRU in said metaverse, based on said most probable path of said at least one vulnerable subject VRU, a three-dimensional volume of relevance S' that is used to filter out elements and / or events that are of relevance for said at least one vulnerable subject VRU, as they are contained within said three-dimensional volume of relevance, said three- dimensional volume of relevance S' that is associated with the digital twin of said at least one vulnerable subject being a volume that extends along said determined most probable path of said at least one vulnerable subject VRU and that has a cross-section dimensioned to include therewithin the cross-section of said at least one vulnerable subject VRU, detect, in said metaverse, any risk of intersection between the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volume of relevance of said at least one vulnerable subject VRU, and when a risk of intersection is detected, send a feedback message from said remote server C to one or more vehicles and to said at least one vulnerable subject VRU to signal that a risk of intersection has been detected.

[0109] In solutions according to the present description, said personal electronic device 17 of said at least one vulnerable subject VRU may include a position sensor configured to determine the current geographical position of said at least one vulnerable subject VRU.

[0110] Said personal electronic device 17 of said at least one vulnerable subject VRU may thus be configured to continuously transmit data related to the geographical position of said at least one vulnerable subject VRU to said remote server C directly and / or via electronic controllers of autonomous driving vehicles or intelligent infrastructures located within a determined distance from said at least one vulnerable subject VRU.

[0111] In this case, the remote server C may be configured to determine the most probable path of said at least one vulnerable subject VRU, not only on the basis of the information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU, but also on the basis of the data related to the geographical position of said at least one vulnerable subject VRU.

[0112] Moreover, in solutions according to the present description, the personal electronic device 17 of said at least one vulnerable subject VRU may be configured to continue to transmit dynamically to said remote server C, directly and / or via electronic controllers of autonomous driving vehicles or infrastructures F located within a determined distance from said at least one vulnerable subject VRU, the information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU, for the repetition by the remote server C of said operations that the remote server C performs continuously and cyclically.

[0113] Similarly, in solutions according to the present description, the electronic controller 3 of each autonomous driving vehicle may be configured to continue to transmit dynamically to said remote server C the data related to the geographical position of the vehicle, the operative conditions of the vehicle including data related to a state and dynamics of the vehicle, and the information indicating the cognitive state, intentions, and emotions of the driver for the repetition by the remote server C of said operations that the remote server C performs continuously and cyclically.

[0114] Figures 6 and 7 of the attached drawings show an example of a real scenario where multiple autonomous driving vehicles are moving, highlighting the advantages of the present solution. Both the figures display the scenario in both a plan view (i.e., in a horizontal xy plane) and an elevation view (i.e., in a vertical xz plane). Both Figures 6 and 7 refer to the same real situation, which involves two overlapping roads: a lower road S1 and an upper road S2 superimposed on road S1. The lower road S1 is traveled by vehicles V11 , V12, V13, V14. The upper road S2 is traveled by vehicles V21 , V22, V23 moving along the left lane (with reference to the direction of travel) of road S2, and by a vehicle V24 traveling along the right lane of road S2. Finally, an ambulance A is approaching on the left lane of the upper road S2.

[0115] With reference to Figure 6, in conventional solutions, the control system for autonomous driving vehicles is based on a two-dimensional representation of the road network, such as the plan view in the upper part of Figure 6. In this representation, all vehicles traveling on the lower road S1 and the upper road S2 appear to be on the same road. Therefore, if the system is configured to alert all vehicles on the path of ambulance A, it will primarily alert vehicle V14 (which is within an influence sphere S associated with ambulance A), and subsequently all vehicles V23, V13, V22, V11 , V21 . In reality, vehicles V14, V13, and V11 do not need to be alerted, as they are on the lower road S1 and do not interfere with the path of ambulance A.

[0116] Figure 7, on the other hand, refers to the case of the present solution. As indicated, the system according to the present solution is configured to perform the filtering operation of elements and events that may interfere with the path of each autonomous driving vehicle based on a three-dimensional virtual simulation (metaverse) of the real situation, including a three- dimensional representation of the road network, which thus allows distinguishing the two roads S1 , S2 at different levels and identifying the vehicles at each level.

[0117] For illustration convenience, Figure 7, like Figure 6, shows a plan view in the xy plane and an elevation view in the xz plane. However, it is to be imagined that in the simulation constructed by the server C, the scenario is represented by a three-dimensional combination of the two views in Figure 7.

[0118] In the metaverse constructed by the remote server C, a digital twin of each vehicle in the real system is created based on data related to the state and dynamics of the vehicles, data related to the geographical position of the vehicles, and information determined via the processing of output signals from the EEG sensors 9 associated with the drivers of these vehicles. Specifically, such remote server C is capable of generating for each of the vehicles a respective three-dimensional volume of relevance based on the data received from the electronic controller 3 of the vehicle. Thus, the system according to the present solution can use a three- dimensional reproduction of the scenario in the metaverse, including the three-dimensional volumes of relevance of each of the digital twins of the vehicles, to filter events and elements that may interfere with the path of each digital twin of every vehicle.

[0119] Assuming that the two views in Figure 7 correspond to the three- dimensional reproduction constructed in the metaverse, this representation identifies the digital twins of the vehicles in the real scenario, indicated respectively as A’, V11 ’, V12’, V13’, V14’, V21 ’, V22’, V23’, V24’ in Figure 7.

[0120] The remote server C continuously receives, from the communication module of each real vehicle, the current geographical position, data related to the state of each vehicle such as speed and dynamics data, and the cognitive state, intentions, and emotions of the driver of each vehicle. The remote server C is thus able to associate, in the metaverse, a position and a most probable path (determined by the remote server C based on the data related to the state and dynamics of the vehicle) to the digital twin of each vehicle and to predict, based on the cognitive state, intentions, and emotions of the driver, possible corrections to this path, such as corrections obtained based on the intention to turn or change lanes, panic conditions making the vehicle's behavior more unpredictable, and / or similar factors, and, consequently, is also able to determine a three-dimensional volume of relevance associated with each digital twin of a vehicle.

[0121] This three-dimensional volume of relevance associated with each digital twin of a vehicle is used to filter elements and / or events that are relevant to the vehicle itself, as they are contained within said three- dimensional volume of relevance. The aforementioned three-dimensional volume of relevance associated with each digital twin of a vehicle is determined by constructing a volume that extends along the most probable path calculated for the vehicle, also considering any corrections to this path obtained based on the cognitive state, intentions, and emotions of the driver, and that has a cross-section sized to include within it the cross-section of the vehicle.

[0122] For example, with reference to Figure 7, to the digital twin of the vehicle ambulance A’ is assigned a three-dimensional volume of relevance S’ consisting of a cylinder extending in front of vehicle A along the most probable path of vehicle A, i.e., along the left lane of road S2, assuming that the information determined through the processing of output signals from the EEG sensors 9 associated with the ambulance driver does not indicate a contrary intention, and that has a circular cross-section containing therewithin the cross-section of the digital twin of the vehicle. To this end, it can be expected that the messages sent by each real vehicle to server C also contain data related to the vehicle's characteristics, such as the type of vehicle (car, bus, truck, etc.) and the length and width of the vehicle.

[0123] In the system according to the present solution, the filtering operation of vehicles interfering with the vehicle ambulance A’ is thus performed in the metaverse, exploiting the three-dimensional representation of the road network, which therefore allows filtering as vehicles actually on the path of the digital twin of the vehicle ambulance A’ only the digital twins of vehicles V21 ’, V22’, and V23’ (if the information determined through the processing of output signals from the EEG sensors 9 associated with the drivers of these vehicles indicates an intention to maintain the same lane), while excluding the digital twins of vehicles on the lower road ST or on a different lane along the upper road S2’ (such as vehicle V24’), unless the information determined through the processing of output signals from the EEG sensors 9 associated with the drivers of vehicles on a different lane of the same upper road S2’ indicates an intention to change lanes or an imminent lane change toward that of ambulance A’.

[0124] Once the filtering operation described above is performed, the system can be configured to send a feedback message from the remote server C to the communication modules T of the affected vehicles (in the example, V21 , V22, V23) containing data related to a modified path for each vehicle, used to limit or prevent the previously detected intersection of vehicles so as to avoid dangerous situations, for example, to have each of these vehicles stop at the side of the road (as schematically represented in the plan view of Figure 7).

[0125] Figure 8 schematically illustrates another real scenario where a two- dimensional representation (in this case in the vertical plane) of the road network can give rise to the identification of false interferences, which is instead avoided in the case of the present solution. Figure 8 specifically refers to a highway with two parallel carriageways S1 , S2 traveled by vehicles in opposite directions (from left to right, with reference to the drawing, carriageway S1 , and from right to left, with reference to the drawing, carriageway S2). Each of the two carriageways S1 , S2 includes two parallel lanes. Along carriageway S1 , vehicles V11 , V12, V13 are present in the right lane and a vehicle V14 in the left lane. On carriageway S2, vehicles V21 , V22, V23 are present in the left lane and a vehicle V24 in the right lane. Moreover, an ambulance vehicle A is approaching along the left lane of carriageway S2.

[0126] An example of intervention of the system according to the present solution to avoid a collision risk is illustrated in Figure 9. This figure shows a plan view and an elevation view of a two-lane road traveled by vehicles all moving in the same direction. Figure 9 aims to illustrate the three- dimensional reproduction of the scenario reconstructed by the server C, which should therefore be understood as a combination of the plan view and elevation view shown in Figure 9. In such three-dimensional reproduction is placed the digital twin HV' of a reference vehicle HV ("Host Vehicle") traveling on the left lane. In the reproduction of the real scenario are also placed the digital twins RO' of multiple remote objects RO ("Remote Objects"), some of which are the digital twins ROT, RO2', RO3' of other autonomous vehicles RO1 , RO2, RO3, while one is the digital twin RO4' of a drone RO4.

[0127] The remote server C receives from the communication modules of each vehicle messages containing data related to the geographical position, data related to the vehicle state such as speed and dynamics data, and information determined by processing the output signals from the EEG sensors 9 associated with each vehicle.

[0128] With particular reference to the reference vehicle HV and the vehicle closest to it, namely RO1 , the server C processes the data related to their positions, their most probable paths (determined by the remote server C based on data related to the state and dynamics of the vehicle), and the information determined by processing the output signals from the EEG sensors 9, in order to construct a three-dimensional volume of relevance associated with the digital twin of each of such vehicles.

[0129] In the case of vehicle HV, considering the cognitive state, intentions, and emotions of the driver, two alternative most probable paths are predicted. A first most probable path of vehicle HV consists of continuing along the left lane. A second most probable path of vehicle HV involves shifting to the right lane. To account for both possibilities, in the metaverse constituting the simulation of the real scenario, a three-dimensional volume of relevance SHV' is constructed for the digital twin HV', extending ahead of HV' along the road, progressively widening to encompass both lanes.

[0130] In the case of vehicle RO1 , the system detects a speed higher than that of vehicle RO2 ahead of RO1 , so a shift of RO1 to the left lane to overtake vehicle RO2 is predicted, unless the information determined through processing the output signals from the EEG sensors 9 associated with the driver of vehicle RO1 indicates an intention to slow down. In the metaverse constituting the simulation of the real scenario, a three- dimensional volume of relevance SROT is thus constructed for the digital twin ROT, extending ahead of ROT along the predicted path.

[0131] The processing performed by the remote server C based on the above-described simulation of the real scenario leads to predicting a collision between vehicles HV and RO1 in the area indicated with I in Figure 9.

[0132] Consequently, the server C can send a feedback signal to vehicles HV and RO1 to avoid a collision. The feedback signal may, for example, contain data related to a modified path for vehicle RO1 involving remaining in the right lane at reduced speed and / or data related to a modified path for vehicle HV, for example, consisting of remaining in the left lane at reduced speed until shifting to the right lane once it is clear. The processing of data related to the modified paths is preferably performed using artificial intelligence algorithms.

[0133] The electronic controllers 3 of each vehicle update the data related to the vehicle state such as speed and dynamics based on said feedback signal and continue to transmit messages containing the updated data to remote server C.

[0134] Figure 10 of the attached drawings shows an example of a real scenario where a vehicle Vi and a vulnerable subject VRU are moving, thus referring to solutions where the control of the plurality of autonomous vehicles is also performed based on communications exchanged, via remote server C, between autonomous vehicles and vulnerable subjects VRU located adjacent to or along a road, outside the autonomous vehicles.

[0135] Figure 10 shows exclusively a plan view, i.e. , in a horizontal xy plane, and, also in this case, when referring to the simulation of the real scenario of Figure 10 constructed by the server C, a simulated scenario corresponding to a three-dimensional representation of the real scenario shown in plan view is obtained.

[0136] Figure 10 illustrates two roads, a first road SA, traveled by an autonomous vehicle Vi, and a second road SB, traveled by a vulnerable subject VRU, for example, a pedestrian.

[0137] In a first exemplary scenario, the two roads illustrated in Figure 10, namely, the first road SA and the second road SB, are considered as two overlapping roads. For example, road SB may correspond to a bridge overlapping road SA.

[0138] In the metaverse constructed by remote server C, a digital twin of vehicle Vi of the real system is created based on the most probable path of such vehicle Vi (determined by the remote server C based on data related to the state and dynamics of vehicle Vi), data related to the geographical position of the vehicle, and information determined by processing the output signals from the EEG sensors 9 associated with the driver of such vehicle Vi. In particular, such remote server C is capable of generating a three- dimensional volume of relevance for vehicle Vi based on the data received from the electronic controller 3 of vehicle Vi.

[0139] In particular, such remote server C may be configured to: determine, based on the data related to the operating conditions of vehicle Vi including data related to a state and dynamics of such vehicle Vi, a most probable path ("Most-Probable-Path" - MPP) of such vehicle Vi, correct such most probable path MPP of vehicle Vi based on the information indicating the cognitive state, intentions, and emotions of the driver of such vehicle Vi, obtaining a corrected most probable path, i.e., a most probable path that takes into account the brain activity of the driver of vehicle Vi, create in the metaverse a digital twin of vehicle Vi , associating such digital twin with the corrected most probable path, and create for such digital twin of vehicle Vi, based on the corrected most probable path associated with such digital twin, a three- dimensional volume of relevance S’ that is used to filter elements and / or events that are relevant to such vehicle Vi, as they are contained within such three-dimensional volume of relevance, such three-dimensional volume of relevance S’ being a volume that extends along such corrected most probable path calculated for vehicle Vi and that has a cross-section sized to include therewithin the cross-section of vehicle Vi.

[0140] Furthermore, in the metaverse constructed by the remote server C, a digital twin of the vulnerable subject VRU of the real system is created based on the information determined by processing the output signals from the EEG sensors 9 associated with such vulnerable subject VRU. In particular, such remote server C is capable of generating a three- dimensional volume of relevance for the vulnerable subject VRU based on the data received from the portable electronic device 17 of the vulnerable subject VRU.

[0141] In particular, such remote server C may be configured to: determine, based on such information indicating the cognitive state, intentions, and emotions of the vulnerable subject VRU, a most probable path of such at least one vulnerable subject VRU, create in such metaverse a digital twin of such vulnerable subject VRU, associating such digital twin with the determined most probable path, and create for such digital twin of the vulnerable subject VRU, based on the determined most probable path, a three-dimensional volume of relevance S’ that is used to filter elements and / or events that are relevant to such at least one vulnerable subject VRU, as they are contained within such three-dimensional volume of relevance, such three-dimensional volume of relevance S’ associated with the digital twin of the vulnerable subject being a volume that extends along such determined most probable path and that has a cross-section sized to include therewithin the crosssection of the at least one vulnerable subject VRU.

[0142] Therefore, the system according to the present solution is capable of using said three-dimensional reproduction of the scenario in the metaverse comprising the three-dimensional volumes of relevance of the digital twin of vehicle Vi and of the digital twin of the vulnerable subject VRU to filter events and elements that may interfere with the path of the digital twin of vehicle Vi.

[0143] The system according to the present solution is configured to perform the filtering operation of elements and events that may interfere with the path of said autonomous driving vehicle Vi based on a three-dimensional virtual simulation (metaverse) of the real situation, including a three- dimensional representation of the road network, which thus allows distinguishing the two roads SA and SB located at different levels and detecting that vehicle Vi and the vulnerable subject VRU are at different levels.

[0144] Since the remote server C continuously receives: the current geographical position, data related to the state of vehicle Vi, and data related to the dynamics of said vehicle Vi, as well as the cognitive state, intentions, and emotions of the driver of said vehicle from the communication module of the real vehicle Vi, and the cognitive state, intentions, and emotions of the vulnerable subject VRU and, optionally, a current geographical position of said vulnerable subject, from the portable electronic device 17 of the vulnerable subject VRU, said remote server C is capable of continuously associating, in the metaverse, a position and a simulated most probable path (determined by the remote server C based on the data relating to the state and dynamics of said vehicle Vi and, optionally, corrected based on the cognitive state, intentions, and emotions of the driver of vehicle Vi) to the digital twin of vehicle Vi and a position and a simulated most probable path to the vulnerable subject VRU (determined based on the information relating to the brain data of the vulnerable subject VRU received and, optionally, the position of said VRU subject), therefore, consequently, it is also capable of continuously determining a three-dimensional volume of relevance associated with the digital twin of vehicle Vi and a three-dimensional volume of relevance associated with the digital twin of the vulnerable subject VRU.

[0145] The aforementioned three-dimensional relevance volumes associated with vehicle Vi and the vulnerable subject VRU are determined by constructing a volume that extends along the most probable path, also taking into account any corrections to said path obtained based on the cognitive state, intentions, and emotions, and have a cross-section sized to include therewithin the cross-section of vehicle Vi and the vulnerable subject VRU, respectively.

[0146] Such three-dimensional relevance volume associated with the digital twin of vehicle Vi and such three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU are used to filter elements and / or events that are relevant to vehicle Vi itself, as they are contained within said three-dimensional relevance volume, therefore, it is verified whether the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU intersects that associated with the digital twin of vehicle Vi to check for a dangerous situation.

[0147] In the system according to the present solution, the filtering operation of the interfering vehicles and vulnerable subjects is performed in the metaverse, exploiting the three-dimensional representation of the road network which thus allows detecting that vehicle Vi traveling on the first road SA and the vulnerable subject VRU traveling on the second road SB, for example, a bridge overlapping the first road SA, are not interfering regardless of the paths followed by said vehicle Vi and by said vulnerable subject VRU.

[0148] Therefore, no feedback message is sent from the remote server C to vehicle Vi and / or to the vulnerable subject VRU.

[0149] In a second exemplary scenario, the two roads illustrated in Figure 10, that is, the first road SA and the second road SB, are considered as two roads belonging to the same plane, that is, as roads that intersect. Also in this case, similarly to what was previously described, a digital twin of vehicle Vi and a digital twin of the vulnerable subject VRU are created in the metaverse constructed by the remote server C. In particular, as previously described, said remote server C is capable of generating a three-dimensional relevance volume of vehicle Vi and a three-dimensional relevance volume of the vulnerable subject VRU.

[0150] Therefore, the system according to the present solution is capable of using the three-dimensional reproduction of the scenario in the metaverse comprising the three-dimensional relevance volumes of the digital twin of vehicle Vi and of the digital twin of the vulnerable subject VRU to filter events and elements that may interfere with the path of the digital twin of vehicle Vi.

[0151] The system according to the invention is configured to perform the filtering operation of elements and events that may interfere with the path of said autonomous driving vehicle Vi based on a three-dimensional virtual simulation (metaverse) of the real situation, including a three-dimensional representation of the road network, which thus allows detecting that the two roads SA and SB are located at the same level and, therefore, intersect as shown in Figure 10.

[0152] Note that even if roads SA and SB are located at the same level, differently from the scenario previously described wherein the two roads SA and SB were located at different levels, what was described for the scenario with roads at different levels remains valid for the features that regard: the data that are continuously received by the remote server C, both from the communication module of the real vehicle Vi and from the portable electronic device 17 of the vulnerable subject VRU; the operations performed by the remote server C aimed at determining the three-dimensional relevance volume associated with the digital twin of vehicle Vi and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU; the descriptions relating to the determinations of said three- dimensional relevance volumes associated with vehicle Vi and the vulnerable subject VRU.

[0153] Indeed, also in this case the three-dimensional relevance volume associated with the digital twin of vehicle Vi and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU are used to filter elements and / or events that are relevant to vehicle Vi itself, as they are contained within said three-dimensional relevance volume. Therefore, also in this case it is verified whether the three- dimensional relevance volume associated with the digital twin of the vulnerable subject VRU intersects that associated with the digital twin of vehicle Vi to check for a dangerous situation.

[0154] In the system according to the present solution, the filtering operation of interfering vehicles and vulnerable subjects is performed in the metaverse, exploiting the three-dimensional representation of the road network which thus allows detecting that: if the three-dimensional relevance volume associated with the digital twin of vehicle Vi follows, for example, the path indicated with reference PVi in Figure 10, such three-dimensional relevance volume associated with the digital twin of vehicle Vi (traveling on the first road SA) and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU (traveling on the second road SB that intersects the first road SA) are interfering, regardless of the path followed by the vulnerable subject VRU; if the three-dimensional relevance volume associated with the digital twin of vehicle Vi follows, for example, the path indicated with reference PV2 in Figure 10, such three-dimensional relevance volume associated with the digital twin of vehicle Vi (traveling on the first road SA) and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU (traveling on the second road SB that intersects the first road SA) are interfering if the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU follows, for example, the path indicated with reference PS1 in Figure 10; and the three-dimensional relevance volume associated with the digital twin of vehicle Vi (traveling on the first road SA) and the three- dimensional relevance volume associated with the digital twin of the vulnerable subject VRU (traveling on the second road SB intersecting the first road SA) do not interfere in the other cases.

[0155] Note that the cases described above are provided by way of example only, therefore, different scenarios following other paths PV or PS where interference or non-interference occurs between the three-dimensional relevance volume associated with the digital twin of vehicle Vi and the three- dimensional relevance volume associated with the digital twin of the vulnerable subject VRU may also be considered.

[0156] Therefore, in response to having performed the filtering operation described above and not having detected interference between the three- dimensional relevance volume associated with the digital twin of vehicle Vi and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU, no feedback message is sent from the remote server C to vehicle Vi and / or to the vulnerable subject VRU since no dangerous situation exists.

[0157] Conversely, in response to having performed the filtering operation described above and having detected interference between the three- dimensional relevance volume associated with the digital twin of vehicle Vi and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU, the system may be configured to send a feedback message from the remote server C: to the communication module T of vehicle Vi containing data indicating the dangerous situation and suggestions to avoid such situation, for example, by slowing down and letting the vulnerable subject VRU pass first or, if necessary, a message capable of activating an Automatic Emergency Brake (AEB) system of the vehicle; and / or to the portable electronic device 17 of the vulnerable subject VRU containing data indicating the dangerous situation and suggestions to avoid such situation, for example, by prompting the vulnerable subject VRU to pay greater attention.

[0158] Note that, instead of continuously sending the current geographical position of the vulnerable subject VRU to the remote server C via the portable electronic device 17, the presence of such vulnerable subject VRU may be determined using the previously described "GeoNetworking" technique.

[0159] Note that, in embodiments according to Figure 10, the data sent from the portable electronic device 17 to the remote server C may be transmitted either directly or via, for example, an electronic controller of an infrastructure F or the electronic controller 3 of vehicle Vi.

[0160] Note that, in embodiments according to Figure 10, in response to having performed the described filtering operation and having detected interference between the three-dimensional relevance volume associated with the digital twin of vehicle Vi and the three-dimensional relevance volume associated with the digital twin of the vulnerable subject VRU, the previously described feedback messages may be generated, in response to a message indicating the detected interference sent from the remote server C, automatically and autonomously by the electronic controller 3 of vehicle Vi and / or by the portable electronic device 17 of the vulnerable subject VRU.

[0161] Note that conventional solutions such as, for example, the previously described "GeoNetworking" technique are only capable of determining the position of elements included in a valid detection area around vehicle Vi, for example, a circular area of a given radius centered around vehicle Vi, on a two-dimensional plane.

[0162] Therefore, the vulnerable subject VRU would not be detected until entering such detection area, delaying the moment when a dangerous situation is detected. Moreover, false positives due to the presence of overlapping roads such as, for example, bridges or parallel roads cannot be avoided due to the two-dimensional nature of the "GeoNetworking" technique, creating dangerous situations in case of false alerts.

[0163] Figure 11 is a schematic illustrating the exchange of messages between the communication modules of the vehicles and, if considered, of the vulnerable subjects VRU, and the remote server C in situations according to the present description. Figure 11 shows a reference vehicle HV or Vi and a remote vehicle RV or a vulnerable subject VRU with a portable electronic device 17 along a road S.

[0164] Note that although Figure 11 only illustrates communication between a first vehicle HV having a first communication module THV and a second vehicle RV having a second communication module TRV, communication may similarly occur between the first vehicle HV or Vi having a first communication module THV and a vulnerable subject VRU, for example, a pedestrian, having a portable electronic device 17 configured to perform the communication functions of the communication module TRV. As already described, the communication modules of the two vehicles HV or Vi and RV, here indicated as THV and TRV, exchange messages with the remote server C that builds the three-dimensional reproduction of the real scenario. The communication modules of both vehicles THV and TRV send to the server C CAM and EEGM ("Electroencephalography Message") messages comprising respectively the indication of the geographical position and data related to the state and dynamics of each vehicle, and the cognitive state, intentions, and emotions of the driver of each vehicle.

[0165] Similarly, the portable electronic device 17 of the vulnerable subject VRU also exchanges messages with the remote server C, for example, directly or via an electronic controller of an infrastructure F or an electronic controller 3 of an autonomous vehicle, which builds the three-dimensional reproduction of the real scenario. The portable electronic device 17 of the vulnerable subject VRU sends to the server C EEGM messages comprising indications of the cognitive state, intentions, and emotions of the vulnerable subject VRU, and, possibly, CAM messages comprising the indication of the geographical position of such VRU subject.

[0166] Thus, the remote server C continuously receives:

[0167] CAM messages comprising indications of geographical positions and data relating to the state and dynamics of the vehicle and EEGM ("Electroencephalography Message") messages comprising the cognitive state, intentions, and emotions of the vehicle driver, for each of the autonomous vehicles present in the real scenario; and

[0168] EEGM messages comprising indications of the cognitive state, intentions, and emotions of a vulnerable subject VRU and, possibly, CAM messages comprising indications of the geographical position of such VRU subject, for each of the vulnerable subjects present in the real scenario.

[0169] Therefore, such remote server C may be configured to determine, based on the data relating to the state and dynamics of each vehicle, a Most-Probable-Path (MPP) associated with each vehicle, taking into account for possible corrections of such path the EEGM messages associated with the respective vehicle driver.

[0170] Moreover, a most probable path may also be defined for the vulnerable subjects, based on the information contained in the EEGM messages relating to such VRU subjects.

[0171] As seen, in the metaverse consisting of the three-dimensional reproduction of the real scenario, containing the digital twins HV' or Vi' and RV' of the vehicles and, if considered, the digital twins of vulnerable subjects VRU', the system processes the received data and the most probable paths determined for each vehicle and / or, if considered, for each vulnerable subject, so as to define a three-dimensional relevance volume associated with each vehicle and, if considered, a three-dimensional relevance volume associated with each vulnerable subject.

[0172] The remote server C may consequently identify any events and / or elements of the scenario that pose a risk to the safety of the vehicles and / or vulnerable subjects by verifying the presence of intersections between different three-dimensional relevance volumes. In case a criticality is detected, the system processes and sends to the respective vehicles, preferably using artificial intelligence algorithms or determined rules ("Rule- Based"), the necessary modifications to the most probable path, possibly corrected based on the detected brain activity, associated with each vehicle (obtained by the remote server C based on the data relating to the state and dynamics of each vehicle, possibly corrected based on the cognitive state, intentions, and emotions of the driver) in order to avoid any situation dangerous to the safety of the vehicles, their occupants, and / or the vulnerable subjects.

[0173] Note that, in case a criticality relating to a vulnerable subject is detected, the remote server C may process and send an alert message to the portable electronic device 17 of the vulnerable subject VRU involved in the criticality.

[0174] Therefore, as a result of this processing operation, the "metaverse", that is, the remote server C that has performed the processing operations based on the three-dimensional simulation of the real scenario, sends DENM, CAM, or BSM messages, for example, via V2M ("Vehicle to Metaverse") communication, both to the RV vehicle and to the HV or Vi vehicle, which can therefore receive these messages, modify their respective paths if necessary based on the received messages, and transmit a new CAM message comprising updated position data and data relating to the state and dynamics of each vehicle, as well as a new EEGM message comprising the cognitive state, intentions, and emotions of the driver to the remote server C. Furthermore, this remote server C may send, for example, again via V2M ("Vehicle to Metaverse") communication, DENM, CAM, BSM messages, or a different type of alert message to the portable electronic devices 17 of the VRU vulnerable subjects involved in the detected criticality.

[0175] Note that via this V2M communication, it is possible to exchange the previously described DENM, CAM, and / or BSM messages.

[0176] This V2M ("Vehicle to Metaverse") communication is intended in solutions according to the present description as an extension of the term V2N ("Vehicle-to-Network") communication. Indeed, instead of referring to communication with a generic network, this term V2M refers to communication with the metaverse described herein.

[0177] Therefore, in the light of the above, in solutions according to the present description, the information indicating the cognitive state, intentions, and emotions, for example, of the driver or a vulnerable subject, may be transmitted to such remote server C, for example, by an electronic controller 3 of the respective autonomous driving vehicle or by a portable electronic device 17 of such vulnerable subject, via an electroencephalography message, i.e., the EEGM message, having a JavaScript Object Notation (JSON) format, comprising: a header comprising metadata of such EEGM message; and a payload comprising such information indicating the cognitive state, intentions, and emotions of the driver or the vulnerable subject.

[0178] In embodiments according to the present description, the header may comprise: a unique identifier of the transmitted electroencephalography message, that is, a unique identifier of the EEGM message; a time indication related to a moment at which such cognitive state, intentions, and emotions of the driver or the vulnerable subject were determined; and a message type variable indicative of the fact that the transmitted message is an electroencephalography message, that is, an EEGM message as previously described.

[0179] In embodiments according to the present description, the payload may comprise: a unique identifier of the person from whom such information indicating the cognitive state, intentions, and emotions was determined, that is, a unique identifier of the driver or the vulnerable subject associated with such information; an action type variable indicative of a type of action that such person (the driver or the vulnerable subject) intends to perform, for example, proceeding straight, turning right, turning left, and / or similar; a probability of an actual execution of such type of action that the person intends to perform, for example, the probability of actually proceeding straight if an intention to proceed straight is detected for the person, and, preferably, a probability of an actual execution of each type of action comprised in a set of types of actions executable by such person, for example, turning right, turning left, proceeding straight, and / or similar; and a confidence associated with such type of action that the person intends to perform.

[0180] Figures 12 and 13 are block diagrams illustrating an exemplary embodiment of the method according to the present solution, respectively with reference to the operations performed in each autonomous driving vehicle and in each portable electronic device 17 of each vulnerable subject, and the operations performed in the remote server C that builds the metaverse consisting of a three-dimensional simulation of the real scenario.

[0181] With reference to Figure 12, block 100 represents the initial phase of the method, wherein: each electronic controller 3 of the autonomous driving vehicles receives updated data on position, vehicle state and dynamics data (CAM), and the cognitive state, intentions, and emotions of the driver (EEGM); and if the control of the plurality of autonomous driving vehicles is also performed based on communications exchanged between autonomous driving vehicles and VRU vulnerable subjects located adjacent to or along a road, outside of the autonomous vehicles, each portable electronic device 17 of the vulnerable subjects receives updated data on the cognitive state, intentions, and emotions of the subject (EEGM).

[0182] Note that, for a greater clarity of illustration, Figure 12 refers to a vehicle of reference HV ("Host Vehicle"), one or more other vehicles RV ("Remote Vehicle"), and one or more vulnerable subjects VRU.

[0183] Note that if the control of the plurality of autonomous driving vehicles is not performed based on communications exchanged between autonomous driving vehicles and VRU vulnerable subjects, that is, if the vulnerable subjects VRU present in the real scenario are not considered in the scenario simulated in the remote server C, the blocks in Figure 12 relating to such one or more vulnerable subjects VRU are not present.

[0184] Block 101 concerns the start of the operations performed by the electronic controller 3 of the reference HV vehicle, block 102 corresponds to the start of the operations performed in the electronic controller 3 of each of the one or more other RV vehicles, and block 103 corresponds to the start of the operations performed in the portable electronic device 17 of each of the one or more vulnerable subjects.

[0185] The operations performed in the reference HV vehicle are represented by blocks 101 A, 101 B , 101 C , 101 D. The operations performed in the other RV vehicles are represented by blocks 102A, 102B, 102C, 102D. The operations performed in the electronic devices 17 of the vulnerable subjects VRU are represented by blocks 103A, 103B, 103C, 103D.

[0186] In block 101 A, the communication module THV of the reference HV vehicle transmits to the remote server C CAM ("Cooperative Awareness Messages") messages comprising a geographical position and data relating to the state and dynamics of the reference HV vehicle, BSM ("Basic Safety Messages"), and EEGM ("Electroencephalography Message") messages comprising the cognitive state, intentions, and emotions of the driver.

[0187] In block 101 B, the communication module THV of the reference HV vehicle is configured to receive from the remote server C, in case of detected criticalities involving the reference HV vehicle, alert messages, for example, comprising warnings for the driver, and / or messages comprising driving-related modifications, for example, speed changes, path changes, activations of the automatic emergency braking system, or similar, necessary to avoid or mitigate the critical situation.

[0188] In block 101 C, the reference HV vehicle modifies its behavior based on the messages received from the remote server C in block 101 B, for example, activating the automatic emergency braking system, slowing down, stopping, or similar, thus, the path of the reference HV vehicle is modified based on the feedback signals from the remote server C.

[0189] In block 101 D, the method is restarted based on the recalculated and updated data, returning to block 100.

[0190] The operations performed by the electronic controllers 3 of the one or more other RV vehicles are represented by blocks 102A, 102B, 102C, 102D. These operations are the same as those described above with reference to blocks 101A-101 D. In block 102A, the electronic controller 3 of another RV vehicle transmits CAM, BSM, and EEGM messages to the remote server C. In block 102B, alert messages and / or messages comprising path and behavior modifications of the vehicle relating to a dangerous situation involving the RV vehicle are received. In block 102C, the path of the RV vehicle is modified based on feedback signals from the remote server C, following the processing performed in the remote server. In block 102D, the method is restarted based on the modified and updated data, returning to block 100.

[0191] The operations performed in the portable electronic devices 17 of the VRU vulnerable subjects are represented by blocks 103A, 103B, 103C, 103D.

[0192] In block 103A, the portable electronic device 17 of a vulnerable subject VRU transmits to the remote server C EEGM (Electroencephalography Message) messages comprising the cognitive state, intentions, and emotions of the vulnerable subject VRU and, optionally, CAM ("Cooperative Awareness Messages") messages comprising a geographical position of such subject.

[0193] In block 103B, the portable electronic device 17 of the vulnerable subject VRU is configured to receive from the remote server C, in case of detected criticalities involving the vulnerable subject VRU, alert messages, for example, including warnings for the VRU subject.

[0194] In block 101 C, the vulnerable subjects VRU may modify their behavior based on the alert messages received from the remote server C in block 103B, for example, by paying more attention to the road and surrounding vehicles, thus, the path and behavior of the vulnerable subjects VRU may be modified based on the feedback signals from the remote server C. In block 103D, the method is restarted based on the recalculated and updated data, returning to block 100.

[0195] Figure 13 illustrates an example of the operations performed by the remote server C.

[0196] Block 200 represents the start of the method, which includes the creation of a digital twin: for each vehicle (both for the reference vehicle HV and for every other vehicle RV); for each object that is in communication with the remote server; and if the control of the plurality of autonomous driving vehicles is performed based on communications exchanged between autonomous driving vehicles and vulnerable subjects VRU, for each vulnerable subject VRU; based on the data contained in the CAM ("Cooperative Awareness Messages"), BSM ("Basic Safety Messages"), and / or EEGM ("Electroencephalography Message") messages sent by each vehicle and / or object and the data contained in the EEGM ("Electroencephalography Message") messages sent by the portable electronic devices 17 of each vulnerable subject VRU.

[0197] Note that also in this case, if the control of the plurality of autonomous driving vehicles is not performed based on communications exchanged between autonomous driving vehicles and vulnerable subjects VRU, that is, if the vulnerable subjects present in the real scenario are not considered in the three-dimensional simulated reproduction via the remote server C, the blocks and operations described with reference to Figure 13 relating to such one or more vulnerable subjects VRU are not present.

[0198] Following the operation in block 200, the operation in block 201 is performed. In block 201 , the remote server proceeds to determine, based on the data relating to the state and dynamics of each of the vehicles received via CAM messages, a most probable path for each vehicle.

[0199] Note that this most probable path of each vehicle is possibly corrected based on the cognitive state, intentions, and emotions of the respective vehicle driver received via EEGM messages, thus, based on the brain activity of the vehicle driver. Furthermore, in this block 201 , the remote server may proceed to determine, based on the cognitive state, intentions, and emotions of each of the vulnerable subjects VRU received via EEGM messages sent by the portable electronic devices 17, a most probable path for each vulnerable subject.

[0200] Following the operation in block 201 , the operations in blocks 202 and 203 are performed. In block 202, the remote server C proceeds to create and process a three-dimensional relevance volume for each digital twin of a vehicle, object, or vulnerable subject in the simulation created in the remote server, that is, a three-dimensional path ("3D Path") having a corresponding volume of reference, based on the determined most probable paths and the data received via CAM and EEGM messages. In block 203, all the positions of the vehicles, objects, and vulnerable subjects are stored, and the corresponding determined most probable paths and associated events are stored.

[0201] Following the operations in blocks 202 and 203, block 204 is reached, where the analysis procedure is executed, preferably using artificial intelligence algorithms ("Al-Powered") or determined rules ("Rule-Based"), to compute the intersection between the three-dimensional relevance volume of a given vehicle and the events and three-dimensional relevance volumes of other vehicles, objects, and / or vulnerable subjects, as previously stored.

[0202] After the operation in block 204, block 205 is reached, where a classification of the events, positions of vehicles, objects, or vulnerable subjects, and their previously identified potentially interfering most probable paths is performed. Note that this interference is determined, as previously described, by verifying the presence of intersections between the three- dimensional relevance volume of a vehicle and the three-dimensional relevance volumes of other vehicles, objects, and / or vulnerable subjects. Block 206 is then reached, where, using artificial intelligence algorithms or rule-based logic, the collision risk is computed.

[0203] In case of a confirmed risk, in block 207, a feedback message is processed and sent, for example, a DENM ("Decentralized Environmental Notification Message"), a CAM ("Cooperative Awareness Message"), and / or a BSM ("Basic Safety Messages") message with warning functions, for example, via the previously described V2M ("Vehicle-to-Metaverse") communication, to signal one or more dangerous events or to modify the behavior and / or path of the vehicle.

[0204] In block 208, a feedback message DENM and / or BSM is processed and sent to signal positions of dangerous vehicles and / or objects, for example, to a vulnerable subject involved in the intersection.

[0205] These feedback signals can be sent both to the electronic controllers 3 of the vehicles (or objects) involved in the one or more dangerous events and to the portable electronic devices 17 of the vulnerable subjects VRU involved in such one or more dangerous events.

[0206] Following the reception of these feedback signals, the electronic controller of each vehicle and / or object may process a modified path to follow or implement strategies such as, for example, the activation of the automatic emergency braking system, aimed at limiting or preventing such dangerous situations and events.

[0207] Furthermore, vulnerable users, for example, pedestrians, may also modify their behavior based on the warning feedback received via the portable electronic device 17, for example, by paying more attention to the road and the present vehicles.

[0208] As evident from the foregoing description, the system and method according to the present solution allow solving a series of problems connected to conventional "GeoNetworking" methods, thanks to the execution of processing operations in the remote server based on a digital simulation of the real situation that also takes into account data relating to the brain activity of drivers and (possibly) vulnerable subjects. In such digital simulation, a three-dimensional reproduction of the real scenario is used, where the digital twin of each vehicle and, if considered, of each vulnerable subject, is assigned a three-dimensional volume of relevance that defines the space associated with each vehicle and / or vulnerable subject in the continuation of the path of the respective vehicle and / or vulnerable subject.

[0209] The system according to the present solution is immediately adaptable to different environments, urban or rural, and is capable of responding quickly to changes in the road network and the related mobility rules, ensuring a real-time response to dynamic variations of the data. The construction of multidimensional digital twins of the vehicles, as well as of objects, infrastructures, and vulnerable subjects, allows for a better understanding of the correlation between the elements (vehicles, buildings, pedestrians) of the simulation.

[0210] Therefore, the solution described in detail in this document facilitates the realization of a system for communication between and control of a plurality of vehicles equipped with autonomous driving functions or predictive trajectory driving functions and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personal electronic device or drones.

[0211] The electronic controller of each vehicle is configured to continuously transmit data relating to the geographical position and operative conditions of the vehicle, the information indicating the cognitive state, intentions, and emotions of the driver to a remote server that builds a metaverse consisting of a three-dimensional virtual simulation of the scenario in which the autonomous driving vehicles and / or connected objects move, including a three-dimensional representation of the road network, configured to distinguish between road paths that are adjacent to each other and / or that intersect with each other but are located at different heights.

[0212] In said metaverse, a digital twin of each vehicle and each connected object is created, and for each of such vehicles, a respective most probable path is computed and associated therewith, and consequently, a three- dimensional relevance volume used to filter elements and / or events that are relevant to the vehicle as they are contained within the three-dimensional relevance volume of the vehicle. In case a risk of intersection between relevance volumes of different vehicles is detected in the metaverse, a feedback message is sent to one or more vehicles.

[0213] Furthermore, solutions according to the present description refer to a method for communication between and control of a plurality of vehicles equipped with autonomous driving functions or predictive trajectory driving functions and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personal electronic device or drones.

[0214] Each vehicle in such plurality of autonomous driving vehicles includes: a plurality of sensors 4 for detecting the environment outside of the vehicle, a plurality of sensors 5 for detecting the operating conditions of the vehicle, a position sensor 6 for determining the current geographical position of the vehicle, one or more sensors 9 of vital parameters of the vehicle driver, including EEG sensors 9, that is, electroencephalogram sensors, configured to monitor the brain activity of the driver, a vehicle driving system 2, an electronic controller 3 configured to receive output signals from such sensors and to issue commands to such vehicle driving system 2, and a communication module T, connected to such electronic controller 3 and configured to communicate with such remote server C and with the communication modules T of other autonomous driving vehicles traveling in the vicinity of the vehicle and / or with personal electronic devices of vulnerable subjects present in the vicinity of the vehicle.

[0215] The method according to the present description comprises: determining, via the electronic controller 3 of the autonomous driving vehicle and using the output signals from the EEG sensors 9, the following information: the cognitive state of the driver, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the driver, and the emotions of the driver, including in particular a state of sudden fear due to a perception of danger, continuously transmitting, via the electronic controller 3 of each autonomous driving vehicle, data relating to the geographical position of the vehicle, data relating to the operative conditions of the vehicle including data relating to a state and dynamics of the vehicle, and the information indicating the cognitive state, intentions, and emotions of the driver to such remote server C, building, via such remote server C, a metaverse consisting of a three-dimensional virtual simulation of a scenario including the autonomous driving vehicles and such objects connected to such remote server, including a three-dimensional representation of a road network, configured so as to enable distinguishing between road paths that are adjacent to each other and / or that intersect with each other but are located at different heights, and further including a representation of infrastructures and / or objects located along and adjacent to the road network.

[0216] Moreover, the method according to the present description comprises performing, via the remote server C and in a continuous and cyclic manner, the following operations: receiving from the electronic controller 3 of each autonomous driving vehicle such data relating to the geographical position of the vehicle, such data relating to the operative conditions of the vehicle including data relating to a state and dynamics of the vehicle, and such information indicating the cognitive state, intentions, and emotions of the driver, determining, based on the data relating to the operative conditions of the vehicle including data relating to a state and dynamics of the vehicle, a most probable path MPP of each autonomous driving vehicle, correcting such most probable path MPP of each autonomous driving vehicle based on the information indicating the cognitive state, intentions, and emotions of the vehicle driver, obtaining a corrected most probable path, creating in such metaverse a digital twin, for example, the digital twins HV’ and RV’, of each vehicle, for example, the vehicles HV and RV respectively, and / or connected object, associating a respective corrected most probable path with each digital twin of a vehicle, creating, for each digital twin of a vehicle in such metaverse, based on the corrected most probable path associated with each digital twin of a vehicle, a three-dimensional relevance volume S’ that is used to filter out elements and / or events that are of relevance for each vehicle, as they are contained within such three-dimensional relevance volume, such three- dimensional relevance volume S’ that is associated with each digital twin of a vehicle being a volume that extends along such corrected most probable path calculated for the vehicle and that has a cross-section sized to include therewithin the cross-section of the vehicle, detecting, in such metaverse, any risk of intersection between the three-dimensional relevance volume of the digital twin of each vehicle with the three-dimensional relevance volumes of the digital twins of other vehicles or with the representations of infrastructures or objects located along or adjacent to the road network, and when a risk of intersection is detected, sending a feedback message from such remote server C to one or more vehicles or objects to signal that a risk of intersection has been detected, when a risk of intersection is signaled, defining, in the electronic controllers of one or more of such vehicles or objects, actions to reduce or eliminate such detected risk of intersection.

[0217] Furthermore, solutions according to the present description may refer to (if the control of the plurality of autonomous driving vehicles is performed taking into account vulnerable subjects VRU located adjacent to or along a road, outside the autonomous vehicles) a method wherein such one or more objects capable of being connected to the remote server include at least one vulnerable subject VRU located adjacent to or along the road, outside the autonomous driving vehicles, for example, a pedestrian such as a child or elderly person, a cyclist or motorcyclist, or a disabled person in a wheelchair.

[0218] The at least one vulnerable subject VRU may be equipped with a personal electronic device 17 and such method may comprise: providing one or more sensors 9 of vital parameters of such at least one vulnerable subject VRU, including EEG sensors 9 to monitor the brain activity of the at least one vulnerable subject VRU, and processing, via such personal electronic device 17, preferably a smartphone or tablet provided to the at least one vulnerable subject VRU, the output signals from the EEG sensors 9 of the at least one vulnerable subject VRU, determining the following information: the cognitive state of the at least one vulnerable subject VRU, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the at least one vulnerable subject VRU, and the emotions of the at least one vulnerable subject VRU, including in particular a state of sudden fear due to a perception of danger.

[0219] In this case, the method according to the present description may further comprise: continuously transmitting, via such personal electronic device 17 of the at least one vulnerable subject VRU, the information indicating the cognitive state, intentions, and emotions of the at least one vulnerable subject VRU to such remote server C directly and / or via electronic controllers 3 of autonomous driving vehicles or infrastructures F located within a determined distance from such at least one vulnerable subject VRU, performing continuously and cyclically, via the remote server C, the following operations: receiving from the personal electronic device 17 of the at least one vulnerable subject VRU such information indicating the cognitive state, intentions, and emotions of the at least one vulnerable subject VRU, determining, based on such information indicating the cognitive state, intentions, and emotions of the at least one vulnerable subject VRU, a most probable path of such at least one vulnerable subject VRU, creating, in such metaverse, a digital twin of such at least one vulnerable subject VRU, associating it with such most probable path of such at least one vulnerable subject VRU, creating, for such digital twin of such at least one vulnerable subject VRU in such metaverse, based on such most probable path of such at least one vulnerable subject VRU, a three-dimensional relevance volume S’ that is used to filter out elements and / or events that are of relevance for such at least one vulnerable subject VRU, as they are contained within such three- dimensional relevance volume, such three-dimensional relevance volume S’ that is associated with the digital twin of the at least one vulnerable subject being a volume that extends along such determined most probable path of such at least one vulnerable subject VRU and that has a crosssection sized to include therewithin the cross-section of the at least one vulnerable subject VRU, detecting, in such metaverse, any risk of intersection between the three-dimensional relevance volume of the digital twin of each vehicle with the three-dimensional relevance volume of the at least one vulnerable subject VRU, and when a risk of intersection is detected, send a feedback message from said remote server C to one or more vehicles and to said at least one vulnerable subject VRU to signal that a risk of intersection has been detected.

[0220] In methods according to the present description, said personal electronic device 17 of said at least one vulnerable subject VRU may include a position sensor configured to determine the current geographical position of said at least one vulnerable subject VRU.

[0221] In this case, the method may comprise an operation of continuously transmitting, via said personal electronic device 17 of said at least one vulnerable subject VRU, data relating to the geographical position of said at least one vulnerable subject VRU to said remote server C directly and / or via electronic controllers 3 of autonomous driving vehicles or infrastructures F located within a determined distance from said at least one vulnerable subject VRU.

[0222] Moreover, in this case, the operation of determining, performed via the remote server C, based on said information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU, the most probable path of said at least one vulnerable subject VRU may be further performed based on said data relating to the geographical position of said at least one vulnerable subject VRU.

[0223] Additionally, in embodiments of methods according to the present description, the personal electronic device 17 of said at least one vulnerable subject VRU may continue to dynamically transmit to said remote server C, directly and / or via electronic controllers 3 of autonomous driving vehicles or infrastructures F located within a determined distance from said at least one vulnerable subject VRU, the information indicating the cognitive state, intentions, and emotions of said at least one vulnerable subject VRU, for the repetition by the remote server C of said operations that the remote server C performs continuously and cyclically.

[0224] Similarly, in embodiments of methods according to the present description, the electronic controller 3 of each autonomous driving vehicle may continue to dynamically transmit to said remote server C the data relating to the geographical position of the vehicle, the operative conditions of the vehicle comprising data relating to a state and dynamics of the vehicle, and the information indicating the cognitive state, intentions, and emotions of the driver, for the repetition by the remote server C of said operations that the remote server C performs continuously and cyclically. In embodiments of methods according to the present description, the operation of transmitting, for example, performed by the electronic controllers 3 of each of the autonomous driving vehicles or by the portable electronic device 17 of each vulnerable subject VRU, the information indicating the cognitive state, intentions, and emotions, for example, of the vehicle driver or the vulnerable subject having said portable electronic device 17, to the remote server C may be performed via an electroencephalography message, that is, the previously described EEGM message, having a JavaScript Object Notation, JSON, format, comprising: a header field comprising metadata of said message; and a payload field comprising said information indicating the cognitive state, intentions, and emotions of the driver or vulnerable subject.

[0225] In embodiments according to the present description, said header may comprise: a unique identifier of the transmitted electroencephalography message; a time indication related to a moment at which said cognitive state, said intentions, and said emotions of the driver or vulnerable subject were determined; and a message type variable indicative of the fact that the transmitted message is an electroencephalography message, that is, an EEGM message.

[0226] In embodiments according to the present description, said payload may comprise: a unique identifier of the person from whom said information indicating the cognitive state, intentions, and emotions is determined, that is, a unique identifier of the vehicle driver or vulnerable subject having the portable electronic device 17 associated with said information; an action type variable indicative of a type of action that said person intends to perform, for example, turning right, turning left, proceeding straight, and / or similar; a probability of an actual execution of said type of action that said person intends to perform, for example, the probability of actually turning right if it is detected that the person intends to turn right, and, preferably, a probability of an actual execution of each type of action included in a set of types of action executable by said person comprising, for example, turning right, turning left, proceeding straight, and / or similar; and a confidence associated with said type of action that said person intends to perform.

[0227] Thus, it can be understood how the solution described in the present detailed description may provide one or more of the following advantages: increasing the awareness regarding the intentions of other drivers and / or vulnerable subjects, reducing road accidents due to misinterpretations of intentions; optimization of the bandwidth both in transmission and reception, since only data relating to the vehicle state and data relating to vehicle dynamics are transmitted, and the most probable path is calculated within the remote server; from the perspective of vulnerable users, the low latency between communications may lead to shorter driver reaction times, who respond to dangerous situations, for example, by braking, more quickly; reducing the computational capacity required by each vehicle, since the operations of processing, simulating, and estimating of the three- dimensional relevance volumes and the possible intersections between them are performed within a server external to the vehicle; and the use of three-dimensional relevance volumes, that is, three- dimensional paths, instead of two-dimensional paths allows filtering, for each of the vehicles, only the three-dimensional relevance volumes that are actually relevant to said vehicle, thus, filtering via the server the intentions and messages that are geometrically relevant or not, increasing the overall efficiency of processing, transmissions, and warnings sent to drivers regarding dangerous situations.

[0228] While maintaining the fundamental principles, the details and embodiments may vary, even appreciably, from what has been described, purely by way of example, without departing from the scope of protection.

[0229] The scope of protection is defined by the attached claims.

Claims

1. CLAIMS1. A system for communication between, and control of, a plurality of vehicles having autonomous driving functions or path predictive driving functions, and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personal electronic device or drones, wherein each autonomous driving vehicle includes: a plurality of sensors (4) for detecting the environment outside the vehicle, a plurality of sensors (5) for detecting the operating conditions of the vehicle, a position sensor (6) for determining the current geographical position of the vehicle, one or more sensors (9) of vital parameters of the driver of the vehicle, including EEG sensors (9) configured to monitor the brain activity of the driver, a vehicle driving system (2), an electronic controller (3) configured to receive output signals from said sensors (4, 5, 6, 9) and to send commands to said vehicle driving system (2), and a communication module (T), connected to said electronic controller (3) and configured to communicate with said remote server (C) and the communication modules (T) of other autonomous vehicles traveling in the vicinity of the vehicle and / or with personal electronic devices of vulnerable subjects present in the vicinity of the vehicle, wherein the electronic controller (3) of each autonomous vehicle is configured to process the output signals from the EEG sensors (9) to determine the following information: the cognitive state of the driver, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the driver, and the emotions of the driver, including in particular a state of sudden fear due to a perception of danger, wherein the electronic controller (3) of each autonomousvehicle is configured to continuously transmit data related to the geographical position of the vehicle, data related to operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and the information indicating the cognitive state, the intentions, and the emotions of the driver to said remote server (C), wherein the remote server (C) is configured to build a metaverse consisting of a three-dimensional virtual simulation of a scenario comprising the autonomous driving vehicles and said objects which are connected to said remote server, including a three-dimensional representation of a road network, configured so as to enable distinguishing between road routes that are adjacent to each other and / or that intersect with each other but are located at different heights, and further including a representation of infrastructures and / or objects that are located along and adjacent to the road network, said remote server (C) being further configured to perform, in a continuous and cyclic manner, the following operations: receiving from the electronic controller (3) of each autonomous vehicle said data related to the geographical position of the vehicle, said data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and said information indicating the cognitive state, the intentions, and the emotions of the driver, determining, based on the data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, a most probable path of each autonomous vehicle, correcting said most probable path of each autonomous vehicle based on the information indicating the cognitive state, the intentions, and the emotions of the driver, obtaining a corrected most probable path, creating, in said metaverse, a digital twin (HV1, RV') of each vehicle (HV, RV, Vi) and / or connected object, associating a respective corrected most probable path with each digital twin of a vehicle, creating, for each digital twin of a vehicle in said metaverse, based on the corrected most probable path associated with each digital twin of a vehicle (HV1, RV'), a three-dimensional volume of relevance (S') whichis used to filter out elements and / or events that are of relevance for each vehicle for the reason that it is contained within said three-dimensional volume of relevance, said three-dimensional volume of relevance (S') that is associated with each digital twin of a vehicle being a volume that extends along said corrected most probable path calculated for the vehicle and that has a cross-section dimensioned to include therewithin the cross-section of the vehicle, detecting, in said metaverse, any risk of intersection of the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volume of relevance of the digital twins of other vehicles or with representations of infrastructures or objects located along or adjacent to the road network, and when a risk of intersection is detected, sending a feedback message from said remote server (C) to one or more vehicles (HV, RV) or objects to signal that a risk of intersection has been detected, when a risk of intersection is signalled, defining, in the electronic controllers of one or more of said vehicles (HV, RV) or objects, actions to reduce or eliminate said risk of intersection detected.

2. The system according to claim 1 , wherein said one or more objects capable of being connected to the remote server comprise at least one vulnerable subject (VRU) located adjacent to or along the road, outside the autonomous vehicles, preferably a pedestrian, in particular a child or an elder person, a cyclist or a motorcyclist, or disabled subjects in wheelchairs, said at least one vulnerable subject (VRU) being equipped with a personal electronic device (17), wherein said system further comprises: one or more sensors (9) of vital parameters of said at least one vulnerable subject (VRU), including EEG sensors (9) configured to monitor the brain activity of the at least one vulnerable subject (VRU), and said personal electronic device (17), preferably a smartphone or tablet in the equipment of the at least one vulnerable subject (VRU), configured to process the output signals from the EEG sensors (9) of the at least one vulnerable subject (VRU) to determine the following information: the cognitive state of the at least one vulnerable subject (VRU), with reference to the degree of attention / distraction, sense of fatigue, and degreeof drowsiness, the intentions of the at least one vulnerable subject (VRU), and the emotions of the at least one vulnerable subject (VRU), including in particular a state of sudden fear due to a perception of danger, wherein said personal electronic device (17) of the at least one vulnerable subject (VRU) is configured to continuously transmit the information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU) to said remote server (C) directly and / or via electronic controllers (3) of autonomous vehicles or of infrastructures (F) located within a determined distance from said at least one vulnerable subject (VRU), wherein said operations performed in a continuous and cyclic manner by the remote server (C) further comprise: receiving from the personal electronic device (17) of the at least one vulnerable subject (VRU) said information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), determining, based on said information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), a most probable path of said at least one vulnerable subject (VRU), creating, in said metaverse, a digital twin of said at least one vulnerable subject (VRU) associating it to said most probable path of said at least one vulnerable subject (VRU), creating for said digital twin of said at least one vulnerable subject (VRU) in said metaverse, based on said most probable path of said at least one vulnerable subject (VRU), a three-dimensional volume of relevance (S') which is used to filter out elements and / or events that are of relevance for said at least one vulnerable subject (VRU) for the reason that it is contained within said three-dimensional volume of relevance, said three-dimensional volume of relevance (S') that is associated with the digital twin of the at least one vulnerable subject being a volume that extends along said determined most probable path of said at least one vulnerable subject (VRU) and that has a cross-section dimensioned to include therewithin the cross-section of the at least one vulnerable subject (VRU), detecting, in said metaverse, any risk of intersection of thethree-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volume of relevance of the digital twin of the at least one vulnerable subject (VRU), and when a risk of intersection is detected, sending a feedback message from said remote server (C) to one or more vehicles (HV, RV) and / or to the personal electronic device (17) of the at least one vulnerable subject (VRU) to signal that a risk of intersection has been detected.

3. The system according to claim 2, wherein said personal electronic device (17) of the at least one vulnerable subject (VRU) comprises a position sensor configured to determine the current geographical position of the at least one vulnerable subject (VRU), wherein said personal electronic device (17) of the at least one vulnerable subject (VRU) is configured to continuously transmit data related to the geographical position of the at least one vulnerable subject (VRU) to said remote server (C) directly and / or via electronic controllers (3) of autonomous vehicles or of infrastructures (F) located within a determined distance from said at least one vulnerable subject (VRU), and wherein said operation of determining, based on said information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), the most probable path of said at least one vulnerable subject (VRU) is further performed based on said data related to the geographical position of the at least one vulnerable subject (VRU).

4. The system according to claim 2 or claim 3, wherein the personal electronic device (17) of the at least one vulnerable subject (VRU) is configured to keep on transmitting dynamically to said remote server (C), directly and / or via electronic controllers (3) of autonomous vehicles or of infrastructures (F) located within a determined distance from said at least one vulnerable subject (VRU), the information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), for the repetition by the remote server (C) of said operations which the remote server (C) performs in a continuous and cyclic manner.

5. The system according to any one of the previous claims, wherein the electronic controller (3) of each autonomous vehicle (HV, RV) is configured to keep on transmitting dynamically to said remote server (C) the data related to the geographical position of the vehicle, to the operativeconditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and the information indicating the cognitive state, the intentions, and the emotions of the driver for the repetition by the remote server (C) of said operations which the remote server (C) performs in a continuous and cyclic manner.

6. The system according to any one of the previous claims, wherein said information indicating the cognitive state, the intentions, and the emotions are transmitted to said remote server (C) via an electroencephalography, EEG, message having a JavaScript Object Notation, JSON, format, comprising: a header comprising metadata of said message; and a payload comprising said information indicating the cognitive state, the intentions, and the emotions; preferably wherein said header comprises: a unique identifier of the transmitted EEG message; a time indication related to a moment at which said cognitive state, said intentions, and said emotions were determined; and a message type variable indicative of the fact that the transmitted message is an EEG message; and wherein said payload comprises: a unique identifier of a person, preferably a driver of an autonomous vehicle or a vulnerable subject, associated with said information indicating the cognitive state, the intentions, and the emotions; an action type variable indicative of a type of action that said person wants to perform; a probability of an actual performance of said type of action that said person wants to perform and, preferably, a probability of an actual performance of each type of action comprised in a set of types of actions performable by said person; and a confidence associated with said type of action that said person wants to perform.

7. A method for communication between, and control of, a plurality of vehicles having autonomous driving functions or path predictive driving functions, and / or one or more objects capable of being connected to a remote server, such as vulnerable subjects equipped with a personalelectronic device or drones, wherein each autonomous driving vehicle includes: a plurality of sensors (4) for detecting the environment outside the vehicle, a plurality of sensors (5) for detecting the operating conditions of the vehicle, a position sensor (6) for determining the current geographical position of the vehicle, one or more sensors (9) of vital parameters of the driver of the vehicle, including EEG sensors (9) configured to monitor the brain activity of the driver, a vehicle driving system (2), an electronic controller (3) configured to receive output signals from said sensors (4, 5, 6, 9) and to send commands to said vehicle driving system (2), and a communication module (T), connected to said electronic controller (3) and configured to communicate with said remote server (C) and the communication modules (T) of other autonomous vehicles traveling in the vicinity of the vehicle and / or with personal electronic devices of vulnerable subjects present in the vicinity of the vehicle, wherein, via said electronic controller (3) of each autonomous vehicle, the output signals from the EEG sensors (9) are used to determine the following information: the cognitive state of the driver, with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the driver, and the emotions of the driver, including in particular a state of sudden fear due to a perception of danger, wherein the electronic controller (3) of each autonomous vehicle continuously transmits data related to the geographical position of the vehicle, data related to operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and the information indicating the cognitive state, the intentions, and the emotions of the driver to said remote server (C), wherein in said remote server (C) is built a metaverseconsisting of a three-dimensional virtual simulation of a scenario comprising the autonomous driving vehicles and said objects which are connected to the remote server, including a three-dimensional representation of a road network, configured so as to enable distinguishing between road routes that are adjacent to each other and / or that intersect with each other but are located at different heights, and further including a representation of infrastructures and / or objects that are located along and adjacent to the road network, wherein in said remote server (C) are further performed, in a continuous and cyclic manner, the following operations: receiving from the electronic controller (3) of each autonomous vehicle said data related to the geographical position of the vehicle, said data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and said information indicating the cognitive state, the intentions, and the emotions of the driver, determining, based on the data related to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, a most probable path of each autonomous vehicle, correcting said most probable path of each autonomous vehicle based on the information indicating the cognitive state, the intentions, and the emotions of the driver, obtaining a corrected most probable path, creating, in said metaverse, a digital twin (HV1, RV') of each vehicle (HV, RV, Vi) and / or connected object, associating a respective corrected most probable path with each digital twin of a vehicle, creating, for each digital twin of a vehicle in said metaverse, based on the corrected most probable path associated with each digital twin of a vehicle (HV1, RV'), a three-dimensional volume of relevance (S') which is used to filter out elements and / or events that are of relevance for each vehicle for the reason that it is contained within said three-dimensional volume of relevance, said three-dimensional volume of relevance (S') that is associated with each digital twin of a vehicle being a volume that extends along said corrected most probable path calculated for the vehicle and that has a cross-section dimensioned to include therewithin the cross-section ofthe vehicle, detecting, in said metaverse, any risk of intersection of the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volumes of relevance of the digital twins of other vehicles or with representations of infrastructures or objects located along or adjacent to the road network, and when a risk of intersection is detected, sending a feedback message from said remote server (C) to one or more vehicles (HV, RV) or objects to signal that a risk of intersection has been detected, when a risk of intersection is signalled, defining, in the electronic controllers of one or more of said vehicles (HV, RV) or objects, actions to reduce or eliminate said risk of intersection detected.

8. The method according to claim 7, wherein said one or more objects capable of being connected to the remote server comprise at least one vulnerable subject (VRU) located adjacent to or along the road, outside the autonomous vehicles, preferably a pedestrian, in particular a child or an elder person, a cyclist or a motorcyclist, or disabled subjects in wheelchairs, said at least one vulnerable subject (VRU) being equipped with a personal electronic device (17), wherein said method comprises: providing one or more sensors (9) of vital parameters of said at least one vulnerable subject (VRU), including EEG sensors (9) for monitoring the brain activity of the at least one vulnerable subject (VRU), and processing, via said personal electronic device (17), preferably a smartphone or tablet in the equipment of the at least one vulnerable subject (VRU), the output signals from the EEG sensors (9) of the at least one vulnerable subject (VRU), determining the following information: the cognitive state of the at least one vulnerable subject (VRU), with reference to the degree of attention / distraction, sense of fatigue, and degree of drowsiness, the intentions of the at least one vulnerable subject (VRU), and the emotions of the at least one vulnerable subject (VRU), including in particular a state of sudden fear due to a perception of danger,wherein said method further comprises: continuously transmitting, via said personal electronic device (17) of the at least one vulnerable subject (VRU), the information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU) to said remote server (C) directly and / or via electronic controllers (3) of autonomous vehicles or of infrastructures (F) located within a determined distance from said at least one vulnerable subject (VRU), performing in a continuous and cyclic manner, via the remote server (C), the following operations: receiving from the personal electronic device (17) of the at least one vulnerable subject (VRU) said information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), determining, based on said information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), a most probable path of said at least one vulnerable subject (VRU), creating, in said metaverse, a digital twin of said at least one vulnerable subject (VRU) associating it to said most probable path of said at least one vulnerable subject (VRU), creating, for said digital twin of said at least one vulnerable subject (VRU) in said metaverse, based on said most probable path of said at least one vulnerable subject (VRU), a three-dimensional volume of relevance (S') which is used to filter out elements and / or events that are of relevance for said at least one vulnerable subject (VRU) for the reason that it is contained within said three-dimensional volume of relevance, said three-dimensional volume of relevance (S') that is associated with the digital twin of the at least one vulnerable subject being a volume that extends along said determined most probable path of said at least one vulnerable subject (VRU) and that has a cross-section dimensioned to include therewithin the cross-section of the at least one vulnerable subject (VRU), detecting, in said metaverse, any risk of intersection of the three-dimensional volume of relevance of the digital twin of each vehicle with the three-dimensional volume of relevance of the digital twin of the at least one vulnerable subject (VRU), and when a risk of intersection is detected, sending a feedbackmessage from said remote server (C) to one or more vehicles (HV, RV) and / or to the personal electronic device (17) of the at least one vulnerable subject (VRU) to signal that a risk of intersection has been detected.

9. The method according to claim 8, wherein said personal electronic device (17) of the at least one vulnerable subject (VRU) comprises a position sensor configured to determine the current geographical position of the at least one vulnerable subject (VRU), wherein said method comprises continuously transmitting, via said personal electronic device (17) of the at least one vulnerable subject (VRU), data related to the geographical position of the at least one vulnerable subject (VRU) to said remote server (C) directly and / or via electronic controllers (3) of autonomous vehicles or of infrastructures (F) located within a determined distance from said at least one vulnerable subject (VRU), and wherein said operation of determining, based on said information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), the most probable path of said at least one vulnerable subject (VRU) is further performed based on said data related to the geographical position of the at least one vulnerable subject (VRU).

10. The method according to claim 8 or claim 9, wherein the personal electronic device (17) of the at least one vulnerable subject (VRU) keeps on transmitting dynamically to said remote server (C), directly and / or via electronic controllers (3) of autonomous vehicles or of infrastructures (F) located within a determined distance from said at least one vulnerable subject (VRU), the information indicating the cognitive state, the intentions, and the emotions of the at least one vulnerable subject (VRU), for the repetition by the remote server (C) of said operations which the remote server (C) performs in a continuous and cyclic manner.

11. The method according to any one of claims 7 to 10, wherein the electronic controller (3) of each autonomous vehicle (HV, RV) keeps on transmitting dynamically to said remote server (C) the data related to the geographical position of the vehicle, to the operative conditions of the vehicle comprising data related to a state and to dynamics of the vehicle, and the information indicating the cognitive state, the intentions, and the emotions of the driver for the repetition by the remote server (C) of said operations which the remote server (C) performs in a continuous and cyclicmanner.

12. The method according to any one of the claims 7 to 11 , wherein said operation of transmitting said information indicating the cognitive state, the intentions, and the emotions to said remote server (C) is performed via an electroencephalography, EEG, message having a JavaScript Object Notation, JSON, format, comprising: a header comprising metadata of said message; and a payload comprising said information indicating the cognitive state, the intentions, and the emotions; preferably wherein said header comprises: a unique identifier of the transmitted EEG message; a time indication related to a moment at which said cognitive state, said intentions, and said emotions were determined; and a message type variable indicative of the fact that the transmitted message is an EEG message; and wherein said payload comprises: a unique identifier of a person, preferably a driver of an autonomous vehicle or a vulnerable subject, associated with said information indicating the cognitive state, the intentions, and the emotions; an action type variable indicative of a type of action that said person wants to perform; a probability of an actual performance of said type of action that said person wants to perform and, preferably, a probability of an actual performance of each type of action comprised in a set of types of actions performable by said person; and a confidence associated with said type of action that said person wants to perform.

Images