System and method for automated inspection of equipment of a public transportation network

Abstract

A system (100) for realizing an automated inspection of equipment (1) of a public transportation system, characterized in that it comprises at least: - a control system (10) comprising at least one electronic processing unit (12) configured to elaborate an inspection plan for an equipment (1) to be inspected; - at least one drone (30) which is configured to execute said inspection plan, wherein said at least one drone (30) is provided with at least one device (32, 34) adapted to collect data and / or one or more images / videos of said equipment (1) while executing said inspection plan; and wherein at least one of said at least one electronic processing unit (12) and a computer system (35) onboard of the at least one drone (30) is configured to allow identifying an abnormality in the equipment (1) inspected by comparing the collected data and / or one or more images / videos with respective data and / or one or more corresponding images indicative of a reference condition of said equipment (1).

Description

[0001] The present invention relates to a system and a method for realizing automated inspections of equipment of a public transportation network.

[0002] The present invention is particularly suitable for realizing automated inspections of railways vehicles or components of a railway network and will be described by making reference to such application, without intending in any way to limit its possible application to other types of public transportation networks, such as bus, metro or tram transportation networks.

[0003] Inspections of equipment, for example of locomotives or components of a railway network, are important in order to detect any anomaly as early as possible, the onset of any anomaly that may affect not only the proper operation of a specific piece of equipment but in some cases also the entire network in which it is used.

[0004] Nowadays, one solution for carrying out such inspection activities foresees the intervention of operators who perform direct visual inspections, for example of a locomotive once parked in a depot, and / or use handheld devices to take pictures to be analyzed later on.

[0005] This solution is rather expensive because it requires the intervention of one or more trained operators, is time consuming, and in any case chances of human errors are not negligible.

[0006] Another solution foresees the use of cameras which are installed in fixed inspection positions and capture images of passing trains.

[0007] This solution entails the realization of a substantial number of fixed inspection stations, and vehicles needs to be passed through these inspection points. Further, the use of this solution is usually limited to inspecting passing vehicles and only some exterior parts thereof exposed to the cameras.

[0008] The present invention is aimed at facing such issues, and in particular at providing a solution for inspecting equipment, such as vehicles or components, of a transportation network which is more effective, simpler and easier to be used compared with known solutions.

[0009] This aim is achieved by a system for realizing an automated inspection of equipment of a public transportation system, characterized in that it comprises at least: a control system comprising at least one electronic processing unit configured to elaborate an inspection plan for an equipment to be inspected; at least one drone which is configured to execute said inspection plan, wherein said at least one drone is provided with at least one device adapted to collect data and / or one or more images / videos of said equipment while executing said inspection plan; and wherein at least one of said at least one electronic processing unit and a computer system onboard of the at least one drone is configured to allow identifying an abnormality in the equipment inspected by comparing the collected data and / or one or more images / videos with respective data and / or one or more corresponding images indicative of a reference condition of said equipment.

[0010] This aim is also achieved by a method for realizing an automated inspection of equipment of a public transportation system, characterized in that it comprises at least the following steps: (a): elaborating, by means of at least one of said at least one electronic processing unit and a computer system onboard of the drone, an inspection plan for an equipment to be inspected; (b): executing, by means of one or more drones said inspection plan, wherein said one or more drones are provided with at least one device adapted to collect data and / or one or more images / videos of said equipment while executing said inspection plan; (c): identifying an abnormality in the equipment inspected, by comparing, via at least one of said at least one electronic processing unit and a computer system onboard of the at least one drone, the collected data and / or one or more images with respective data and / or one or more corresponding images indicative of a reference condition of said equipment.

[0011] Further characteristics and advantages will become apparent from the description of some preferred but not exclusive exemplary embodiments of a system and method according to the present disclosure, illustrated only by way of non-limitative examples with the accompanying drawings, wherein: Figure 1 is a block diagram schematically illustrating a system for realizing automated inspections of equipment of a public transport network, according to the present invention; Figure 2 is a flow diagram schematically illustrating a system for realizing automated inspections of equipment of a public transport network, according to the present invention;

[0012] It should be noted that in order to clearly and concisely describe the present disclosure, the drawings may not necessarily be to scale and certain features of the disclosure may be shown in somewhat schematic form.

[0013] Further, when the term "adapted" or "arranged" or "configured" or "shaped", is used herein while referring to any component as a whole, or to any part of a component, or to a combination of components, it has to be understood that it means and encompasses correspondingly either the structure, and / or configuration and / or form and / or positioning of the related component or part thereof, or combinations, such term refers to.

[0014] In particular, for electronic and / or software means, each of the above listed terms means and encompasses electronic circuits or parts thereof, as well as stored, embedded or running software codes and / or routines, algorithms, or complete programs, suitably designed for achieving the technical result and / or the functional performances for which such means are devised.

[0015] Figures 1 and 2 schematically illustrate a system and a method, indicated by the corresponding overall reference numbers 100 and 200, for realizing automated inspections of equipment of a public transportation network.

[0016] For example, the piece of equipment to be inspected can be a public transportation vehicle, such as a railway vehicle, be it a single car, a locomotive, parts of or an entire convoy, or a component of a transportation network, for instance signals, a catenary, switch points, wayside devices, point machines, balises, et cetera.

[0017] An example of such equipment is schematically represented in figure 1 in the form of an icon representing a train 1; in the following; for ease of illustration, reference will be made to the train 1 as a representative example of any possible piece of equipment to be inspected.

[0018] The system 100 comprises at least: a control system 10 which comprises at least one electronic processing unit 12 configured to elaborate an inspection plan for an equipment to be inspected, for instance the train 1; at least one drone 30 which is configured to execute the inspection plan once elaborated by the electronic processing unit 12.

[0019] In one possible example, when the equipment 1 to be inspected is a public transportation vehicle, such as a train, the inspection plan for an equipment to be inspected is for example elaborated based on real time location / movement information and schedule information of the public transport vehicle.

[0020] The at least one drone 30 is provided with at least one device adapted to collect data and / or one or more images / videos of the train 1 while executing the inspection plan.

[0021] The at least one drone 30 is also provided with an on-board computer system, schematically represented in figure 1 by the reference number 35, which is adapted to process the collected data and / or images / videos collected by the at least one device.

[0022] Usefully, in the system 100 according to the invention, at least one of the at least one electronic processing unit 12 and the computer system 35 onboard of the drone 30 is configured to allow identifying a possible abnormality in the train inspected by comparing the collected data and / or one or more images / videos with respective data and / or one or more corresponding images / videos indicative of a reference condition of said equipment, for instance the train 1.

[0023] Preferably, both the at least one electronic processing unit 12 and the computer system 35 onboard of the drone 30 are configured to allow identifying, jointly or independently from each other, a possible abnormality in the train inspected by comparing the collected data and / or one or more images / videos with respective data and / or one or more corresponding images / videos indicative of a reference condition of said equipment, for instance the train 1.

[0024] To this end, the electronic processing unit 12 and / or the drone's on board computer 35 runs suitable and per se known machine vision and deep learning algorithms.

[0025] When at least one abnormality is identified, the at least one electronic processing unit 12 and / or the onboard computer system 35 is configured to issue a warning signal, e.g. to the operator 3.

[0026] The reference condition is a condition that is considered normal or anyhow acceptable for a proper / acceptable functioning of the equipment inspected.

[0027] The reference condition of said equipment, for instance the train 1 or all the relevant data required to identify the abnormalities can be loaded from the control system 10 on the drone 30 along with the inspection plan before starting the inspection.

[0028] In this way, it is possible to detect as early as possible the occurrence or onset of abnormalities, such as structural misalignments between parts, abnormalities due to mechanical fatigues, structural wear and tear or corrosion of parts, liquid / gas leakages, et cetera.

[0029] Clearly, in the system 100 according to the invention there could be used a plurality of drones 30, each provided with a respective device adapted to collect data and / or one or more images / videos of the train 1 while executing the inspection plan, and a respective onboard computer system 35 adapted to collect data and / or one or more images / videos of the train 1.

[0030] In such case, the electronic processing unit 12 can elaborate only one inspection plan with tasks assigned to the various drones 30 to be executed independently and / or in coordination among each other.

[0031] Alternatively, each drone 30 may receive from the control system 10 a specific inspection plan.

[0032] The electronic processing unit 12 can be part of or associated with a centralized server 13 of the control system 10.

[0033] Further, as illustrated in the exemplary embodiment of figure 1, the control system 10, which can be part of or constituted by the automated train supervision system (ATS) of a railway network, comprises for instance a database 14, a communication system 15, e.g. wireless, and an operations control center 17.

[0034] The control system 10 can comprise also a drone control system 18 for supervision and / or remote manual control of the drone(s) 30 used for inspection.

[0035] For instance, the centralized server 13 is provided with all relevant information related to equipment that should be inspected along with train movements and train schedules related information.

[0036] The centralized server 13 can be provided also with all relevant information related to drone 30 and other drones related associated systems 50, 60.

[0037] For example, when the train 1 arrives at the final destination point or a train depot, a scheduler module of the centralized server 13, and in particular the electronic processing unit 12, runs a customized scheduling algorithm to schedule the inspection request for the train 1.

[0038] Then, the scheduled information, namely the inspection plan to be executed, is sent to assigned drone(s) 30 and operator(s) 3 over a secure wireless communication channel for execution.

[0039] The database 14 stores all relevant operational information of the entire inspection process, such as for example in the case of a train 1, but not limited to: train / locomotive information, e.g. car number, type, image information et cetera; waypoints information (geolocation information); survey information, e.g. RFID, waypoint information for all Point of Interests (POI) inside an inspection area; battery charging station information, for instance availability, RFID, waypoints et cetera; drone station information, for example availability, RFID, waypoint et cetera; all scheduling requests assigned by the centralized server 13; inventory of drones and spare parts; maintenance records of drones; scheduling requests for inspection activities.

[0040] Clearly, the various components of the control system 10 may be installed in a common location, e.g. a control room, or in different locations remote from each other.

[0041] According to one possible embodiment, the at least one drone 30 is further provided with at least one removable memory, schematically represented in figure 1 by the reference number 31, which is adapted to store thereon data and / or images / videos collected by the at least one device.

[0042] In this way, when for example a drone 30 is executing the inspection plan, e.g. in an off-line manner, namely it is not in operative communication with the control system 10, the collected data and / or image(s) can be stored on the memory 31 and downloaded and analyzed later on by means of the electronic processing unit 12.

[0043] A drone 30 can also auto transmit all the collected data and / or image(s) from the removable memory 31 to the control system 10 as soon as a secure connection is established with the control system 10.

[0044] Alternatively, when for example a drone 30 is executing the inspection plan in an on-line manner, namely it is in operative communication with the control system 10, the collected data and / or image(s) can be transmitted directly to the control system 10, and in particular sent in input to the electronic processing unit 12 for further elaboration in real time.

[0045] To this end, the or each drone 30 is provided with a wireless communication module 33; further, according to solutions per se known and therefore not described herein in details, the or each drone 30 may comprise one or more of one or more sensors, a lighting source, battery(ies), and a navigation system.

[0046] In one possible embodiment, the at least one device comprises at least one camera 32 adapted to capture one or more images / videos of the train 1 under inspection.

[0047] Accordingly, the at least one electronic processing unit 12 and / or the computer system 35 on board of a drone 30 is configured to identify an abnormality in the train 1 by comparing the one or more images / videos captured with one or more corresponding reference images indicative of a reference condition of the train 1 itself (or of any part thereof).

[0048] For instance, the at least one camera 32 can comprise one or more visual and / or infrared cameras. When a drone executes the inspection plan assigned to it, the camera(s) 32 can record on the memory 31 and / or transmit to the remote control system 10 pictures or videos captured while flying and / or process the collected data using its on-board computer system 35.

[0049] In another possible embodiment, the at least one device comprises at least one sensor 34 adapted to collect data related to the train 1 to be inspected.

[0050] For instance, depending on the applications and / or the inspection requirements, the at least one sensor 34 can comprise one or more ultrasonic sensors, gas sensors, smoke detector sensor, LIDAR et cetera and / or any other sensor which can be mounted on drone for any required inspection operation.

[0051] According to this embodiment, the at least one electronic processing unit 12 and / or the on-board computer system 35 is configured to identify an abnormality in the train 1 inspected by comparing the data captured with one or more corresponding reference data indicative of a reference condition of the train 1 (or of any part thereof).

[0052] Clearly, it has to be understood that in the system 100 according to the invention, both camera(s) and sensor(s) can be used at the same time.

[0053] The data and / or video / images collected, whether initially stored in the memory 31 or transmitted to the control system 10 directly on-line, can be saved in the database 14.

[0054] In one possible embodiment, the at least one electronic processing unit 12 is configured to modify in real time the inspection plan elaborated for an equipment 1 to be inspected while the at least one drone 30 is executing the elaborated inspection plan.

[0055] It is also possible, for whatever reason, that an operator 3 at the drone control center 18 overrides the automatic control of a drone 30 and maneuvers it manually during execution of the assigned inspection plan.

[0056] The remote control of any drone 30 used can be executed for instance via a console 19 or even via a portable electronic device, for instance via an App installed on a mobile phone 20.

[0057] Usefully, the system 100 according to the invention, inspection plans can be generated according to different scenarios or needs, other than the one above mentioned, namely based on the normal on-arrival time of the train 1 at a depot where inspections can be executed.

[0058] For instance, according to a possible embodiment, the control system 10, and in particular the at least one electronic processing unit 12, is configured to elaborate the relevant inspection plan based on at least one of: a time-based inspection schedule for the equipment to be inspected 1; an event related to the equipment 1 to be inspected; a condition-based inspection need for the equipment 1; a break-down of the equipment 1 or part thereof; usage of the equipment 1; regulatory requirements for the equipment 1.

[0059] Inspection plans generated based on a time-base schedule for a piece of equipment, such as a train 1, are the most frequently used inspection triggers. If an equipment is scheduled for inspection on a predetermined schedule, such as every month, or every week, or every day, or every trip, then the inspection plan is executed according to such maintenance schedule.

[0060] Event based inspection triggers mainly depends on some kind of other events. Thus, if such an event happens, it triggers that kind of inspection.

[0061] Condition based situations may trigger an inspection plan as well. In practice, when a certain element of an equipment is not working the way it is supposed to, it could mean something is about to happen and that situation results in triggering a condition-based inspection plan.

[0062] Similarly, when a piece of equipment breaks down and can not be used anymore, this event triggers the elaboration and execution of an inspection plan as well for the failed equipment or for the related equipment.

[0063] Also, when an equipment has operated at a certain output, in the system 100 according to the invention this triggers the elaboration and execution of a usage-based inspection plan.

[0064] Further, inspection plans may be triggered by regulatory requirements. These kind of inspection requests can be considered as high priority requests.

[0065] The control system 10, and in particular its electronic processing unit 12 can be triggered for elaboration and execution of inspection requests automatically, or it can receive the inspection trigger requests as external inputs, e.g. from third-parties (systems or software). The control system 10 can process third party inspection requests based on a defined priority.

[0066] In one possible embodiment, the at least one electronic processing unit 12 is configured to determine if another piece of equipment is in a position unacceptable for the at least one drone 30 to execute the inspection plan of the equipment 1 to be inspected and in the affirmative case to inhibit or postpone the execution of the inspection plan.

[0067] In particular, in case of a train 1 or parts thereof to be inspected, the control system 10, and in particular its electronic processing unit 12 takes into account the real-time location of other trains on parallel tracks along with the schedule information of the train 1 for creating the inspection plan for the drones 30. In practice, the control system 10 may determine whether other trains are on the parallel tracks or not based on train schedules and real time location of trains or by a method in which drones 30 and trains communicate with each other and share the location and movement information. In such scenario, the electronic processing unit 12 can modify the inspection plan for the target train 1, inhibiting or at least postponing its execution.

[0068] In this way, drones 30 can execute the inspection operations in a safer manner.

[0069] In a possible embodiment, the system 100 further comprises one or more sensors 40 for detecting actual weather conditions around the inspection areas, or the control system 10 is in operative communication with an external weather forecast service 42 for receiving information about actual and forecast weather conditions.

[0070] Accordingly, the at least one electronic processing unit 12 is configured to determine when the inspection plan can be executed based on the detected weather conditions or the received information about actual and forecast weather conditions.

[0071] Figure 2 illustrates a method for realizing an automated inspection of equipment, of a public transportation system, such as the train 1, which can be used in connection with or can be executed by means of the components of the above described system 100.

[0072] As illustrated, the method 200 comprises at least the following steps: 210:elaborating, by means of at least one electronic processing unit, such as the electronic processing unit 12, an inspection plan for an equipment, e.g. the train 1, to be inspected;220:executing, by means of one or more drones 30 said inspection plan, wherein said one or more drones 30 are provided with at least one device 32, 34 adapted to collect data and / or one or more images / videos of said equipment 1 while executing said inspection plan;230:identifying an abnormality in the equipment 1 inspected, by comparing, via at least one of said at least one electronic processing unit 12 and a computer system 35 onboard of the drone 30, the collected data and / or one or more images / videos with respective data and / or one or more corresponding images indicative of a reference condition of said equipment 1.

[0073] The method 200 can execute in term of steps all functionalities / tasks performed by the various components of the system 100 as previously described and which are not replicated hereinafter in terms of step features for the sake of conciseness.

[0074] For example, in one possible embodiment, when the equipment 1 is a public transportation vehicle, such as a train, the inspection plan can be elaborated based on real time location / movement information and schedule information of the public transport vehicle.

[0075] In practice, when the system 100 is installed, the drone(s) 30 to be used can execute one or more survey phases. In each phase, operators perform for instance the survey of an entire terminal station area and yard area to find the Point of Interests. POIs are objects or entities which either need to be inspected or which are needed for navigation of drones 30. The survey phase is preferably a one-time process to capture all required information for autonomous inspection operations using drones. For an outdoor navigation survey, operators will manually fly the drones and identify the waypoints for movement of drones in the outdoor yard area. This information is used for outdoor navigation of drones during the execution phase as detailed hereinafter. In this phase, the entire yard or the required area can be surveyed and information collected is saved in the database 14. RFID tags (or equivalent alternative means) mounted on drone charging station(s), drone station(s) and track side components (POls) are also captured during this survey process and saved in the database 14. All points of interest (POls) components will be geotagged and will use RFID tags or equivalent means which will be used during inspection.

[0076] The entire survey process can be repeated multiple times to build a robust waypoint infrastructure for movement of drones 30. An operator can record the following information that can be used later on in the execution phase for inspection activity: waypoints information for navigation inside the yard / depot area (Intermediate geolocations for autonomous movement of drones); waypoints information for navigation of following POIs: drone charging stations, one example of which is schematically represented in figure 1 by the reference number 50; drone station(s), one example of which is schematically represented in figure 1 by the reference number 60; yard POI objects / entities which need to be inspected, for instance like signals, catenary, switch points, wayside devices, point machines, et cetera; train / locomotive waypoint information (dedicated parking location in the yard or train can send parked locations to the control system 10 before end of trip). RFID tag information or equivalent information using alternate technology for following POI items: drone charging station(s); drone station(s); yard POls objects / entities which need to be inspected (like signal, switch points, wayside devices et cetera); trains / locomotive IDs: prior information about all POIs (visual image information and / or infrared information, type of POI or other default inspection information for all the Yard POls objects / entities and trains)

[0077] Other inspection information can be the threshold values for any parameter which needs to be inspected.

[0078] During the execution phase of an inspection plan for example of a train 1, the waypoints information captured during the survey phase are used for instance to automate the inspection activity by the control system 10. For instance, the control system 10 uses the real time train movement and schedule information to automate the inspection activities using the available drones 30. The control system 10 has train information including the number of train cars, train car type, platform information or train geolocation once the train arrives in a yard or in a terminal station. Using this real-time information, the control system 10 fetches the required train related information from database 14 and generates a flight plan for drone navigation. In general, a drone flight plan includes the following information based on the equipment under inspection: waypoints information (e.g. all navigation information for movement of each drone in the yard area); train car waypoint information (e.g. current parking location); train car identifier information; train car type information; yard POI waypoint information ; yard POI identifier information ; yard POI type information ; train information (like prior visual / infrared image information or prior recorded parameters information, CAD model information or any other equivalent information et cetera); yard POI information (like visual / infrared image information or prior recorded parameters information, et cetera); drone(s) ID and any other drone related information; drone charging station or drone station information; assigned operator / supervisor information; inspection method (fully autonomous or remotely controlled).

[0079] The control system 10 can assign the multiple inspection activities to a single drone 30 to efficiently manage the available drones and planned inspection activities. In the case where multiple activities are assigned to a drone 30, the drone 30 can navigate to a further point of interest (POI) without returning to drone station(s) 60 based on the available battery status.

[0080] The control system 10 can use a predictive algorithm to analyze the required inspection time and available battery status to assign the new activity to a drone 30 in real-time.

[0081] Alternatively the control system 10 can also use a predictive algorithm to analyze the required inspection time and available battery status to stop the current inspection activity assigned to the first drone 30 and re-assign the pending inspection activity to a different drone 30, or pause the current inspection activity and resume once the sufficient battery is available with the first drone 30 after the first drone is charged.

[0082] If the inspection activities of drone(s) 30 are completed or paused, drone(s) 30 navigate back to a drone station 60 or a drone charging station 50.

[0083] A drone station 60 is designed to accommodate landing and takeoff of drones which are used in the yard or terminal station area. The drone station(s) may have one or more landing locations and one or more deployment locations to accommodate multiple drones, which may used for various activities like sanitization, inspection et cetera.

[0084] The drone station 60 may be used to fulfill many commands from the control system 10 each day using multiple drones. Therefore, each drone 30 can be configured to support many assigned tasks in the priority based approach. A drone station 60 provides required infrastructure to use different payloads on the same drones for the assigned task. The process to modify the payloads inside the drone station 60 can be manual or automated. The drone station 60 will include services to charge batteries of the drone(s) 30, inspect and / or service the drone(s) 30, and / or perform any other operations related to drone(s) 30.

[0085] The drone charging station(s) 50 can be incorporated into existing structures such as light poles, power poles, signal poles and catenary poles. A drone charging station 50 can also comprise standalone structures to provide additional services like waypoint information for drone 30 navigations in the yard area using RFID tags or equivalent alternative method.

[0086] A drone charging station 50 is designed to accommodate landing and takeoff of drone(s) 30 and there can be multiple drone charging stations positioned inside the yard or terminal station area in such a way that it will help the drone(s) 30 for charging as well as for navigations.

[0087] The control system 10 can run various diagnostics algorithms to decide a drone health and availability of the same drone 30 for next activities.

[0088] The control system 10 will also provide the required waypoint information about the return path to drones whenever the inspection activity is assigned to drones.

[0089] The control system 10 can completely control the activity and can command the drone system 30 to stop the activity at any stage of the inspection process.

[0090] During the inspection operation, each drone 30 communicates the real-time battery status and location to the control system 10, e.g. periodically, where operator(s) 3 can monitor the battery status and location status and act if any parameter goes below a predefined threshold value.

[0091] Each drone charging station(s) 50 and drone stations 60 will communicate the drone health status directly with the control system 10 or using drone 30 wireless channel. The health status may include the real-time battery status, location information or any other parameters to the control system 10 periodically where operator(s) 3 can monitor the drone health.

[0092] In the system 100 and method 200 according to the invention, the inspection plan can be executed in a fully autonomous way or in a remotely controlled one.

[0093] In the fully autonomous inspection way, the control system 10 schedules and assigns all the inspection activities for drones and operators using for instance the information captured in the survey phase. Drone(s) 30 will receive all the required information as a part of an inspection plan elaborated.

[0094] The information of the inspection plan are used by drones 30 for navigation and to perform the inspection tasks. Once the inspection task is completed, each drone 30 can return back to the assigned drone station 60 or to the drone charging station 50. Operators may supervise multiple inspection drones 30 from the operations control center 16. In case of any unmanaged incidents, operator 3 can take over navigation control or drone 30 can return to the nearest drone station 60 or drone charging station 50.

[0095] The homing signals from a homing signal generator and transmitted by the drone 30 can be used by the drone station 60 as a more precise indication of the position of the drone 30 than the GPS position for guiding the drone 30 into the drone station 60.

[0096] In alternative embodiments, a homing signal or other location indicators for the drone station 60 may be transmitted to and received by each drone 30 and used by return guidance software stored at the drone memory to guide the inspection drone 30 into the drone station 60 or drone charging station 50.

[0097] In the remotely controlled modality, the navigation of each drone 30 is remotely controlled by operators 3. The control system 10 assigns the drone(s) and related inspection task to one or more corresponding operators 3 and shares the related drone inspection plan information with operators 3.

[0098] In this remotely controlled modality, drone operators 3 can follow the navigation path provided by the control system 10 or can manually fly the drone 30 to assigned POIs which need to be inspected. Once the inspection process is completed, each operator 3 will have to navigate the controlled drone 30 either back to nearest drone station 60 or drone charging station 50, or navigate them to the next equipment which needs to be inspected.

[0099] In practice, the navigation of drones 30 is a three-step process, namely a first step of outdoor navigation till POI which need to be inspected (navigation from drone station 60 or drone charging station 50 to assigned train car entry point), a second step of navigation for inspection activities (actual inspection procedure may vary as per the type of POI), and a third step of outdoor navigation from a POI which needs to be inspected, to assigned drone station 60 or drone charging station 50.

[0100] For instance, in case of a train 1 to be inspected, the inspection activities of drones 30 are divided for instance in the following steps depending on the train car structure: train set under chassis inspection, train set interior inspection, train set body inspection, train set roof inspection.

[0101] For other trackside POls which are on the track or near the tracks, the inspection activities can vary depending on the structure of the POI or based on the inspection objectives. Few examples of trackside POI inspection points are: catenary cables that can be inspected for example for loosening of their tension by measuring height; switch conditions that can be inspected for instance for debris obstructing switch movements; splice plate or hold-down clip or balise.

[0102] As above indicated, the system 100 and method 200 according to the invention use machine vision-based software and operate for instance according to four operational steps, namely: an image acquisition step, wherein for example visual and IR images / videos are captured and the videos are converted into multiple images. As the drone 30 moves, the camera(s) 32 simultaneously record(s) in the visual and / or infrared images to digital media and / or transmits them to the remote control system 10; after, in an image processing step, images are processed and combined to create both visual and IR of the entire train 1, and then the train images are separated into individual equipment / component. The visual and infrared images are processed for instance frame by frame for any correction using different image processing algorithms and filters. Multiple images are algorithmically stitched together to create unified image of the entire train which will then be used by various image processing algorithms. Recorded image information are used to identify the features and each feature would be identified to determine if it has been recorded before during previous inspection activities; then, in an image inspection step, all identified features are verified against the last known template or reference images to inspect for any changes that might indicate emerging failures and / or for damaged parts and / or for missing components and / or unexpected components. If no previous recording of the subject piece of equipment / component is available, a template of an identical piece in known good condition is used as a reference to be compared, or a computer-generated CAD model is used for comparison. Those areas or individual / components that do not match well with either the reference template or the other similar components are marked as anomalies requiring further investigation. Both visual image and IR images are used by image processing algorithms to identify any anomalies. The majority of anomalies can be detected using two-dimensional imaging, but machine vision applications can capture / create and use 3D images also to detect anomalies which are not possible using 2D images. 3D images or models created using the captured data can be used to identify any anomalies with better accuracy and precision; finally, in a report generation step, for instance the software used generates a detailed report of the detected anomalies. The report generated can include all the related information about the identified anomalies and planned action items to mitigate risks related to the identified anomaly(ies).

[0103] For instance, each equipment / component can be inspected for one or more of the following types of inspection: presence inspection that can be used mainly to check the quantity or the presence / absence of equipment / component on a target object which is under inspection. The proposed solution is capable of differentiating target types as well as counting targets present inside the object which is under inspection; appearance or surface inspection that is mainly used to check for foreign objects detection, any surface modification and any surface contamination with liquid and oil spill on the surface of equipment / components under inspection; a dimension inspection that is a type of appearance inspection and plays an important role to confirm whether components or equipment are in proper shape and size or not according to defined specifications. With image processing, various dimensions can be obtained from captured images. It is easy to measure the dimensions of various components based on this data and to judge whether they are within tolerances; positioning / alignment inspection process for which the image processing can be used for verifying component positioning with respect to equipment; structural inspection where for example 3D image information can be used to inspect the target objects which are under inspection for any change in structure.

[0104] Clearly, the detailed inspection for each component can vary based on the type of equipment, and according to the system and method according to the invention is possible, in principle, to implement in principle any type of inspection need.

[0105] Hence, it is evident from the foregoing description that the system 100 and method 200 according to the present invention allow achieving the intended aim since they allow executing inspection plans and identifying any possible anomaly in a very effective, simple and easy way.

[0106] The system 100 and method 200 thus conceived are susceptible of modifications and variations, all of which are within the scope of the inventive concept as defined in particular by the appended claims; for example, in relation to the specific application, some of the components, e.g. of parts or the whole control system 100 can be positioned remotely, or there could be more than one control unit for sharing in coordination the various tasks and functionalities beforehand described as executed by the electronic processing unit 12.

[0107] All the details may furthermore be replaced with technically equivalent elements.

Claims

1. A system (100) for realizing an automated inspection of equipment (1) of a public transportation system, characterized in that it comprises at least: - a control system (10) comprising at least one electronic processing unit (12) configured to elaborate an inspection plan for an equipment (1) to be inspected; - at least one drone (30) which is configured to execute said inspection plan, wherein said at least one drone (30) is provided with at least one device (32, 34) adapted to collect data and / or one or more images / videos of said equipment (1) while executing said inspection plan; and wherein at least one of said at least one electronic processing unit (12) and a computer system (35) onboard of the at least one drone (30) is configured to allow identifying an abnormality in the equipment (1) inspected by comparing the collected data and / or one or more images / videos with respective data and / or one or more corresponding images indicative of a reference condition of said equipment (1).

2. The system (100) as in claim 1, wherein the at least one electronic processing unit (12) is configured to directly identify an abnormality in the equipment (1) and to issue a warning signal.

3. The system (100) as in claim 1 or 2, wherein said computer system (35) onboard of the at least one drone (30) is configured to directly identify an abnormality in the equipment (1) and to issue a warning signal when the at least one drone (30) is in operative communication with the control system (10).

4. The system (100) as in any one of the preceding claims, wherein the at least one drone (30) is further provided with at least one removable memory (31) adapted to store thereon data and / or images / videos collected by said at least one device (32, 34).

5. The system (100) as in any of the preceding claims, wherein the at least one device (32, 34) comprises at least one camera (32) adapted to capture said one or more images / videos of said equipment (1) to be inspected, and wherein at least one of the at least one electronic processing unit (12) and said computer system (35) onboard of the at least one drone (30) is configured to identify an abnormality in the equipment (1) inspected by comparing the one or more images / videos captured with one or more corresponding reference images indicative of a reference condition of said equipment (1).

6. The system (100) as in any of the preceding claims, wherein the at least one device (32, 34) comprises at least one sensor (34) adapted to collect data related to the equipment (1) to be inspected, and wherein at least one of the at least one electronic processing unit (12) and said computer system (35) onboard of the at least one drone (30) is configured to identify an abnormality in the equipment (1) inspected by comparing the data captured with one or more corresponding reference data indicative of a reference condition of said equipment (1).

7. The system (100) as in any of the preceding claims, wherein the at least one electronic processing unit (12) is configured to modify in real time the inspection plan elaborated for an equipment (1) to be inspected while the at least one drone (30) is executing said elaborated inspection plan.

8. The system (100) as in any of the preceding claims, wherein the at least at least one electronic processing unit (12) is configured to determine if another piece of equipment is in a position unacceptable for the at least one drone (30) to execute the inspection plan of the equipment (1) to be inspected and in the affirmative case to inhibit or postpone the execution of the inspection plan.

9. The system (100) as in any of the preceding claims, wherein the at least one electronic processing unit (12) is configured to elaborate said inspection plan based on at least one of: a time-based inspection schedule for the equipment to be inspected (1); an event related to the equipment (1) to be inspected; a condition-based inspection need for the equipment (1); a break-down of the equipment (1) or part thereof; usage of the equipment (1); regulatory requirements for the equipment (1).

10. The system (100) as in any of the preceding claims, wherein it further comprises one or more sensors (40) for detecting actual weather conditions located within or nearby areas to be inspected, and / or the control system (10) is in operative communication with a weather service (42) for receiving information about actual and forecast weather conditions, and wherein the at least one electronic processing unit (12) is configured to determine when the inspection plan can be executed based on the detected weather conditions or the received information about actual and forecast weather conditions.

11. A method (200) for realizing an automated inspection of equipment (1) of a public transportation system, characterized in that it comprises at least the following steps: (210): elaborating, by means of at least one of said at least one electronic processing unit (12) and a computer system (35) onboard of the drone (30), an inspection plan for an equipment (1) to be inspected; (220): executing, by means of one or more drones (30) said inspection plan, wherein said one or more drones (30) are provided with at least one device (32, 34) adapted to collect data and / or one or more images / videos of said equipment (1) while executing said inspection plan; (230): identifying an abnormality in the equipment (1) inspected, by comparing, via at least one of said at least one electronic processing unit (12) and a computer system (35) onboard of the at least one drone (30), the collected data and / or one or more images with respective data and / or one or more corresponding images indicative of a reference condition of said equipment (1).

Images