Online sleep scoring based on vital signs of a driver of a vehicle

Abstract

A computer program product (7) comprising instructions which, when executed by electronic processing resources (10) of a system (1) for online scoring sleep stages among Awake (W), Light Sleep (LS), Deep Sleep (DS) and REM sleep (R) of a subject, cause the electronic processing resources (10) to: receive (S10-S14) at least one physiological signal of a subject; process (S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject; output (S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject; and determine (S20, S23, S25, S28, S29) the status of the subject based on the outputted score. The computer program product (7) is further configured to, when executed, cause the electronic processing resources (10) to: determine a quantity related to the time needed for a user to resume control of a vehicle (TOT) after being asleep; and provide assistance to the subject based on the quantity related to the time needed for a user to resume control of a vehicle after being asleep (TOT).

Description

[0001] "ONLINE SLEEP SCORING BASED ON VITAL SIGNS OF A DRIVER OF A VEHICLE"'

[0002] Cross-Reference to Related Applications

[0003] This Patent Application claims priority from European Patent Application No. 24213938.4 filed on November 19, 2024 and Italian Patent Application No. 102025000030232 filed on November 18, 2025, the entire disclosure of which is incorporated herein by reference.

[0004] Technical Field of the Invention

[0005] The present invention generally relates to monitoring the status of a driver of a vehicle, in particular to a system for evaluating the status of a driver of a vehicle, more in particular for online scoring sleep stages (in particular, online scoring the levels of sleep) of the driver, based on vital signs of a driver of a vehicle and a related computer program product or software.

[0006] Background of the Invention

[0007] As is known, since there is an overall tendency towards autonomous driving, in Level 4 automation sleeping while driving becomes a possible scenario, including the issue of transitions.

[0008] However, there is a new challenge when the driver is put into a so-called out of the loop-state. Being out of the loop, drivers neither pay attention to the driving task, which is taken over by the system, nor to the system surveillance; if automated systems do not function as they are supposed to, the driver needs to be drawn into the loop as fast as possible or the out of the loop-state needs to be avoided completely. This means that the driver should be drawn back into the loop with the help of an early warning signal when his or her action is needed or is kept in the loop with the help of some kind of dead-man- switch, e.g. touching the steering wheel or buttons now and then. Drivers tend to do more secondary tasks during automated driving, indeed; thus, the out of the loop-state becomes more and more likely.

[0009] It is furthermore noted that a significant factor of an out of the loop-state is sleep, in which awareness and attention are crucially reduced; in particular, sleep and the need to sleep are controlled by external factors, such as day / night cycles, as well as internal factors, so-called endogenous oscillators. Indeed, it is well known that activity level, pulse and body temperature sink during sleep.

[0010] The following concepts and definitions are reported:

[0011] - vigilance, which is defined as the “state or the level of willingness to recognize and react to changes, which occur in random time intervals in a given environment” (from Bergius, 2009, S. 1075);

[0012] - powernaps, which are defined as short sleeping periods that are characterized by the absence of deep sleep, in particular, such powernaps are considered especially beneficial to performance and alertness (from Milner & Cote, 2009); and

[0013] - sleep inertia, which is defined as the “phenomenon of decreased performance and / or disorientation occurring immediately after awakening from sleep relative to presleep status” (from Tassi & Muzet, 2000). In fact, after sleeping, time is needed in order to reach the performance level of being awake. Despite the positive effects of powernaps on performance (as also discussed in Staedt & Riemann, 2007) there is a drop in cognitive performance shortly after waking up, which is referred to as sleep inertia.

[0014] Furthermore, the scoring of sleep stages is one of the essential tasks in sleep analysis. According to the guideline by the American Academy of Sleep Medicine (AASM), sleep scoring consists in dividing sleep into intervals of substantially thirty seconds, called epochs, and in assigning one of the following sleep stages to each epoch:

[0015] - stage W, wakefulness;

[0016] - stage Nl / NREM1 (Non-Rapid Eye Movement 1) (formerly SI);

[0017] - stage N2 / NREM2 (Non-Rapid Eye Movement 2) (formerly S2);

[0018] - stage N3 / NREM3 (Non-Rapid Eye Movement 3) (formerly S3 + S4); and

[0019] - stage R, Rapid eye movement (REM).

[0020] In practice, sleep stages N1-N3 are, in some cases, combined into one NREM stage. Another possible simplification consists in the following stages: Awake, Light Sleep (corresponding to Nl and N2), Deep Sleep (corresponding to N3), and REM sleep (R).

[0021] As is known, sleep scoring can be performed either:

[0022] - manually, i.e. physiological signals are acquired through a dedicated data recording device, and then the recorded signals are analyzed manually by trained sleep experts, scoring the stages of sleep. Mainly electroencephalogram (EEG), Electrooculogram (EOG), and Electromyogram (EMG) signals provide the required input. This approach enables accurate scoring of the sleep stages; however, it also entails a high expenditure of time and financial resources to complete this task. Moreover, despite following the scoring guidelines, manual analysis introduces a certain level of subjectivity, known as interscorer / interrater reliability / variability, leading to the possibility that the results of the evaluation of the same sleep recording by different experts may differ; and

[0023] - automatic, i.e. computer program products or software are used to perform sleep scoring according to suitably defined algorithms that implement the sleep scoring guidelines.

[0024] In further detail, automatic sleep scoring can be performed in different manners, namely:

[0025] - off-line, i.e. the physiological signals are acquired for the entire duration of the sleep, then the recorded signals are processed to identify the sleep stages. Sleep scoring is thus performed after the sleep activity.

[0026] - on-line (also referred to as real-time), i.e. the sleep scoring activity is performed concurrently with the recording of physiological signals. Sleep scoring is thus performed during the sleep activity.

[0027] Off-line and on-line automatic sleep scoring are performed through the analysis either of the electroencephalogram (EEG) signals, or the Photoplethysmography (PPG) signal in combination with motion signals, or Heart Rate (HR) / Heart Rate Variability (HRV) signals in combination with motion signals.

[0028] Typically, the abovementioned sleep scoring is carried out using artificial intelligence techniques such as Artificial Neural Networks (ANN), Support Vector Machines (SVM), Convolutional Neural Networks (CNN) or Long Short-Term Memory (LSM) networks to classify sleep stages.

[0029] Object and Summary of the Invention

[0030] The Applicant notes that human error has been indicated as the causal factor in a large proportion of motor vehicle crashes, with drowsy driving accounting for up to 20% of accidents. Technological advances, such as the introduction of self-driving vehicle capabilities, can potentially eliminate all driving tasks. Vehicles with such capabilities allow the driver to enable and disable the self-driving system but require the driver to be prepared to appropriately respond to a takeover request (TOR) of the vehicle at any time. As the driver’s role shifts from one of active engagement to passive monitoring, partially autonomous systems may introduce new safety concerns.

[0031] Furthermore, the Applicant noted that the currently available techniques and methods are passible of improvements.

[0032] An object of the present invention is thus providing systems and computer program products or software that allow to overcome at least in part the disadvantages of the known prior art.

[0033] According to the present invention, a system for evaluating the status, in particular for online scoring sleep stages, of a driver of a vehicle and a related computer program product or software are provided, as claimed in the appended set of claims.

[0034] Brief Description of the Drawings

[0035] Figure 1 schematically shows a system for online scoring sleep stages of a driver of a vehicle according to the present invention.

[0036] Figures 2-4 schematically show architectures of a system for online scoring sleep stages of a driver of a vehicle according to the present invention.

[0037] Figure 5 shows a flow chart of a method for evaluating the status of a driver of a vehicle implemented by a computer program product loadable and executable by a system for online scoring sleep stages of a driver of a vehicle according to the present invention.

[0038] Description of Preferred Embodiments of the Invention

[0039] The present invention will now be described in detail with reference to the accompanying drawings in order to allow a skilled person to implement it and use it. Various modifications to the described embodiments will be readily apparent to those of skill in the art and the general principles described may be applied to other embodiments and applications without however departing from the protective scope of the present invention as defined in the appended claims. Therefore, the present invention should not be regarded as limited to the embodiments described and illustrated herein but should be allowed the broadest protection scope consistent with the features described and claimed herein. Unless otherwise defined, all technical and scientific terms used herein have the same meaning commonly understood by one of ordinary skill in the art to which the invention belongs. In case of conflict, the present specification, including the definitions provided, will control. Furthermore, the examples are provided for illustrative purposes only and as such should not be considered limiting.

[0040] In particular, the block diagrams included in the attached figures and described below are not to be understood as a representation of the structural features, i.e. constructional limitations, but must be understood as a representation of functional features, i.e. intrinsic properties of the devices defined by the effects obtained, that is to say functional restrictions, which can be implemented in different ways, so as to protect the functionalities thereof (operational capability).

[0041] In order to facilitate the understanding of the embodiments described herein, reference will be made to some specific embodiments and a specific language will be used to describe the same. The terminology used herein is used for the purpose of describing particular embodiments only and is not intended to limit the scope of the present invention.

[0042] To summarise, the present invention relates to a on-line sleep scoring algorithm that is based on analyzing Heart Rate (HR) and / or Heart Rate Variability (HRV) signals in combination with motion signals or Respiration Rate (RR) and / or Respiration Atlas (RA) signals in combination with motion signals. The signals are collected in a predetermined time interval, in particular in 300 seconds frames (without limiting the possibility to run at different frames / second frequency) and processed through a dedicated algorithm, in particular a rule-based algorithm that exploits a simplified model of the autonomous nervous system (ANS). As also reported in further detail below, in contrast with the existing state-of-the-art, the present invention does not entail any artificial intelligence techniques, and it does not require any form of training of the sleep stage classification algorithm.

[0043] The HR and / or HRV signals can be obtained either by means of a Photoplethysmography (PPG) sensor, such as the one provided with wearable devices, a camera-based sensor by means of imaging Photoplethysmography (iPPG) or by means of an Electrocardiogram (ECG) sensor such as the one provided with a chest strap band. In the case of wearable devices, the motion signal is provided e.g. by an inertial platform (such as a 3-axes accelerometer, and possibly 3-axis gyroscope) or the like. In case of a camera, the motion information is obtained by image processing techniques.

[0044] The RR and / or RA signals are obtained by means of devices such as continuous waveform frequency modulated (CW-FM) RADAR device or Ultrawideband (UWB) RADAR. In the case of RADAR device, the motion signal is derived from the RADAR by processing the RADAR range signal (z.e., distance and relative speed between RADAR and target).

[0045] As also disclosed in further detail below, the on-line sleep scoring algorithm provides sleep stage classification as follows:

[0046] - Awake, W;

[0047] - Light Sleep (corresponding to N1 and N2), LS;

[0048] - Deep Sleep (corresponding to N3), DS; and

[0049] - REM sleep, R.

[0050] Moreover, the abovementioned algorithm is configured to identify the time instant of sleep state change. In particular, the abovementioned algorithm is configured to identify:

[0051] - the time the transition between wake and light sleep starts, corresponding to the drowsiness onset;

[0052] - the time the transition between light sleep and deep sleep starts; and

[0053] - the time the transition between deep sleep and REM starts.

[0054] Furthermore, the present invention is configured to implement a SENSE- COMPUTE-ACT mechanism, based on sleep stage classification, aimed at providing online details about the sleep status of the driver, as well as, when required, the estimated takeover transition time.

[0055] Figures 1 and 2 schematically show a system 1 for online scoring sleep stages among Awake (W), Light Sleep (LS), Deep Sleep (DS) and REM sleep (R) of a subject (e.g. a driver of a vehicle) comprising:

[0056] - a sensory system 2, here comprising either a wearable sensor 4 (e.g. a smartwatch), a contactless sensor 5, 6 (e.g. a RADAR, a camera, e.g. an iPPG camera), configured to acquire physiological data of a subject and generate corresponding output signals;

[0057] - electronic computing resources 10, here an embedded platform, configured to communicate with the sensory system 2 to receive the outputted signals (thereby receiving the acquired physiological data) and to store, load and execute, when in use, a computer program product or software or algorithm 7 therein to output data relative to a scoring of the sleep stages of the subject; and

[0058] - an interface to vehicle system 3 configured to communicate with the electronic computing resources 10 and to receive the output generated by the latter to provide feedback to external resources, e.g. further electronic devices external to the system 1 and / or to the subject.

[0059] In particular, as better disclosed in the following paragraphs, the term “online” refers to the fact that the abovementioned system 1 is configured to output a scoring of the sleep stages and act as also disclosed below substantially at the same time at which physiological data is acquired; in other words, the acquisition of the physiological data and the processing and the following steps performed by the present system 1 are real time, meaning that a negligible time interval passes between the acquisition of the physiological data and the further steps performed by the system 1, in particular by the electronic processing resources 10.

[0060] In further detail, considering Figure 2, the sensory system 2 also comprises a sensing unit 12 configured to communicate with the sensors 4, 5, 6, and the electronic computing resources 10, in particular for transmitting to the latter the abovementioned outputted signals. Furthermore, the electronic processing resources 10 here comprises a processing unit 15 configured to receive the data outputted by the sensory system 2 and execute the software 7 to perform the operations disclosed in the following paragraphs. Thus, the sensing unit 12 provides the physical interfaces to the sensors 4, 5, 6 and transfers the physiological data as well as the image to the processing unit 15 which runs the software 7.

[0061] Furthermore, the interface to vehicle system 3 comprises a feedback unit, here comprising either a local feedback unit 17 and / or a remote feedback unit 18, configured to communicate with the electronic computing resources 10 and to receive the output generated by the latter to provide feedback to external resources.

[0062] In particular, the software 7 comprises instructions which, when executed by the electronic processing resources 10, cause the electronic processing resources 10 to:

[0063] - receive (blocks S10-S14) at least one physiological signal of a subject, in particular from the sensory system 2, more in particular from at least one among the wearable sensors 4 and / or the contactless sensors 5, 6;

[0064] - process (blocks S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, namely Awake (W), Light Sleep (LS), Deep Sleep (DS) and REM sleep (R);

[0065] - outputting (blocks S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject; and

[0066] - determine (blocks S20, S23, S25, S28, S29) the status of the subject based on the outputted score.

[0067] Furthermore, the software 7 comprises instructions which, when executed by the electronic processing resources 10, cause the electronic processing resources 10 to:

[0068] - determine a quantity related to the time needed for a user to resume control of a vehicle after being asleep; and

[0069] - provide assistance to the subject based on the quantity related to the time needed for a user to resume control of a vehicle after being asleep.

[0070] In particular, according to an aspect of the present invention, the physiological signals are indicative of vital signs, in particular heart rate HR and / or heart rate variability HRV signals; according to an aspect of the present invention, such vital signs are acquired at a 1 Hz frequency.

[0071] Furthermore, the scoring of the sleep stage is according to the abovementioned four classes, i.e. Awake (W), Light Sleep (LS), Deep Sleep (DS) and REM sleep (R); in addition, the scoring is outputted at the same frequency as the one used for acquiring the vital signs, here 1 Hz.

[0072] Reference is now made to Figure 5, where a flow chart of the method implemented by the software 7, when run by the processing unit 15 and, more in general, by the electronic processing resources 10. It is furthermore noted that the following method is of the iterative type, meaning that the outputted scoring for the sleep state is stored and compared to with the following values to determine the actual status of the subject. Furthermore, the present method is of the online type, meaning that the electronic processing resources 10 are configured to operate in an online manner with the other elements of the system 1.

[0073] In further detail, in order to receive (blocks S10-S14) at least one physiological signal of a subject, the software 7 is configured to cause the electronic processing resources 10 to determine (block S14) if the subject is awake (W) at a first instant i-1, where i is a natural number from 0; in fact, being an iterative process, the software 7 is configured to evaluate at each time instant the status of the subject. Before the abovementioned step, the software 7 is configured to cause the electronic processing resources 10 to:

[0074] - acquire (block S10) from an internal storage the status of the subject at a second time instant i, previous to the first time instant i-1, as well as related indexes;

[0075] - determine (block Sil) if the user stopped the software 7 from working, i.e. the user provided a command to interrupt the operations of the software 7;

[0076] - if the user did not stop the algorithm, acquire (block S12) the vital signals; and

[0077] - define (block S13) one or more reference values.

[0078] If it is determined that the user provided a command to the electronic processing resources 10 to interrupt the software 7 from working, the process is stopped (block S31).

[0079] Furthermore, in order to define (block S13) the one or more reference values, the software 7 is configured to cause the electronic processing resources 10 to:

[0080] - determine a time frame win for observing the subject’s status;

[0081] - determine quantity related to the abovementioned vital signs, namely HR and HRV signals; and

[0082] - determine a standard deviation of the quantity related to the abovementioned vital signs, i.e. of the HR and the HRV signals acquired in the previous steps.

[0083] Furthermore, in order to process (blocks S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software 7 is configured to cause the electronic processing resources 10 to:

[0084] - if the subject is determined to be awake (W) in the first instant i-1, determine (block S15-S18) if one or more values of the vital signs, in particular of the heart rate HR and / or heart rate variability HRV signals, is monotonically reducing for a given amount of observations N, wherein N is a natural number; and

[0085] - if the one or more values of the vital signs, in particular of the heart rate HR and / or heart rate variability HRV signals, are determined to be reducing, determine (block S19) whether in a time frame the one or more vital signs satisfy a stabilization requirement.

[0086] Furthermore, in order to output (blocks S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (blocks S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software 7 is configured to cause the electronic processing resources 10 to, if the one or more vital signs satisfy the stabilization requirement in the time frame, determine (block S20) that the subject is in a light sleep state (LS).

[0087] In particular, the step of determining whether in a time frame the one or more vital signs satisfy the stabilization requirement comprises determining if the difference between the maximum and minimum values of each vital sign is lower than a corresponding threshold value; thus, according to the present invention, the step of determining whether in a time frame the one or more vital signs satisfy the stabilization requirement comprises determining if the difference between the maximum and minimum values of the heart rate HR signal is lower than a heart rate threshold value HRTh.

[0088] Furthermore, the step of determining whether in a time frame the one or more vital signs satisfy the stabilization requirement comprises:

[0089] - determining a further quantity related to the vital signs, here a heart rate smooth HRSmooth value; and

[0090] - determining that, if the quantity related to the vital signs is lower than a related threshold value, here a heart smooth threshold value HRSmoothTh.

[0091] In further detail, the heart rate smooth HRSmooth is a value computed as the moving average of the last m values of the quantity related to the vital signs, here the last m values of the HR signal; here, m isa user-defined constant, for example equal to 120.

[0092] Furthermore, the software 7 is configured to cause the electronic processing resources 10 to, if the subject is determined to be asleep in the first instant i-1, determine if one or more values of the vital signs, in particular of the heart rate HR and / or heart rate variability HRV signals, satisfy one or more requirements for determining which state the subject is.

[0093] In further detail, in order to process (blocks S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software 7 is configured to cause the electronic processing resources 10 to determine (block S21-S22) if the values of the one or more vital signs are above a respective thresholds, e.g. the values of the heart rate HR and / or the heart rate variability HRV signals are above the respective thresholds HRTh and HRVTh

[0094] Furthermore, in order to output (blocks S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (blocks S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software 7 is configured to cause the electronic processing resources 10 to, if the values of the one or more vital signs are determined to be above the respective thresholds, determine (blocks S23) that the subject in the awake state (W).

[0095] In addition, if the values of the one or more vital signs are determined to be below above a respective threshold, namely if at least one among the values of the heart rate HR and the heart rate variability HRV signals are below the respective thresholds HRTh and HRVTh, in order to process (blocks S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software 7 is configured to cause the electronic processing resources 10 to:

[0096] - determine (block S24) one or more statistical quantities related to the further quantity related to the vital signs, i.e. the heart rate smooth HRSmooth value; and

[0097] - determine (block S24) if the values of the one or more statistical quantities related to the further quantity related to the vital signs are below respective threshold values, here HRSmoothTh.

[0098] Furthermore, in order to output (blocks S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (blocks S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software 7 is configured to cause the electronic processing resources 10 to, if it is determined that the values of the one or more statistical quantities related to the further quantity related to the vital signs are below respective threshold values HRSmoothTh, determine (block S25) that the subject is in a deep sleep state (DS).

[0099] In particular, the one or more statistical quantities comprise a standard deviation of further quantity related to the vital signs, i.e. the heart rate smooth HRSmooth value.

[0100] Moreover, if the values of the one or more statistical quantities related to the further quantity related to the vital signs are determined to be above the respective threshold values, here HRSmoothTh, the software 7 is configured to cause the electronic processing resources 10 to implement further checks. In particular, in this case, the one or more statistical quantities also comprise a standard deviation of NN samples, the latter being a measure of heart rate variability HRV signals. Accordingly, in order to process (blocks S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software 7 is configured to cause the electronic processing resources 10 to determine (block S26- S27) if the values of the standard deviation of NN samples are comprised between a first and a second threshold, here SDNNThl, SDNNTh2.

[0101] Furthermore, in order to output (blocks S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (blocks S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software 7 is configured to cause the electronic processing resources 10 to:

[0102] - if the values of the standard deviation of NN samples are comprised between the first and the second threshold SDNNThl, SDNNTh2, determine (block S28) that the subject is in the REM state (R); and

[0103] - if the values of the standard deviation of NN samples are not comprised between the first and the second threshold SDNNThl, SDNNTh2, determine (block S29) that the subject is in the light sleep state (LS).

[0104] It is noted that the abovementioned operations are performed periodically, here every second according to the acquisition of the abovementioned vital signals until the algorithm is stopped. Thus, at the end of every iteration, the related indexes are updated, as well as the scoring value.

[0105] At the end of the abovementioned operations, the system 1 is configured to output the status of the subject, in particular the scoring of the sleep state (if present).

[0106] Based on the current wakefulness, sleep state and transition time, as also anticipated above, in order to determine a quantity related to the time needed for a user to resume control of a vehicle after being asleep, the software 7 is configured to cause the electronic processing resources 10 to determine a quantity related to the ability of the subject to recover, in particular to determine a takeover transition time TOT of the subject. In further detail, the takeover transition time TOT is estimated based on a lookup table that relates the takeover transition time TOT with the subject state, in particular as follows:

[0107] - in case of wakefulness state (awake, W), the takeover transition time TOT is quale to 5 seconds (worst case for adult subject fit for driving, z.e., the driver is not fatigued); - in case of light sleep state (LS), the takeover transition time TOT is 600 seconds (worst case for adult subject fit for driving, z.e., the driver is not fatigued);

[0108] - in case of deep sleep state (DS), the takeover transition time TOT is 1800 seconds (worst case for adult subject fit for driving, z.e., the driver is not fatigued); and

[0109] - in case of REM sleep state (R), the takeover transition time TOT is 900 seconds (worst case for adult subject fit for driving, z.e., the driver is not fatigued).

[0110] The time instant when a transition between one vigilance / sleep state to the next one occurs is provided to enable user-defined actions, i.e. to provide assistance to the user to avoid incurring into a sleep state, namely to minimize the takeover transition time TOT; as soon as the drowsiness onset is identified, feedback is provided to the subject being monitored to avoid the subject entering in the light sleep state (LS).

[0111] As better disclosed in the following, the present system 1 implements a SENSE- COMPUTE-ACT mechanism aimed at providing online details about the sleep status of the driver and, if required, the estimated takeover transition time. In other words, the electronic processing resources 10 are configured to receive biometric input, namely in the form of physiological signals, and classify them according to a number of scoring levels, here four online sleep scoring levels.

[0112] The presently disclosed system 1 can be implemented in different architectural structures, thus expanding the possibility to adopt such a solution in different applications. In particular, from the architectural viewpoint, both integrated and distributed solutions can be implemented, as shown in Figure 3. In further detail, on the left side of the same Figure 3, the system 1 is implemented by comprising an edge device 20 fully integrating different sensors (i.e. the sensory system 2) and the processing unit 15; in this case, the feedback is locally generated. On the right side of the same Figure 3, different sensors (in particular, of the sensory system 2) in a first device 21 locally communicate (e.g., through wireless or wired communication, e.g. through Wi-Fi, BT, RF or the like) with the processing unit 15 being located in a second device 22; in this case, the processing unit 15 generates the feedback.

[0113] Furthermore, according to further aspects of the present invention, both short-range communication as well as the long-range communication for distributed architecture for the system 1 according to the present invention are considered. Specifically, referring to Figure 4, in the short-range solution, a user and the system 1 are in close vicinity as, for instance, the in-cabin automotive applications or a monitoring room in a health centre. Instead, the long-range communication enables new services to the users as well as new loT applications. For this reason, an additional architectural solution for the system 1 is foreseen in the case remote communication is established between different devices, where sensing, processing feedback are independently generated, as shown in Figure 4.

[0114] In particular, considering Figure 4, a first and a second device 30, 31, implementing at least a portion of the system 1 according to the present invention, are configured to communicate with a third device 32 to carry out the method implemented by the software 7. From the disclosure above, the present invention has several advantages.

[0115] In particular, the present invention provides for an online, iterative method, implemented by a related system and computer program product, where the sleep scoring activity is performed concurrently with the recording of physiological signals. Sleep scoring is thus performed during the sleep activity. Consequently, an objective method is proposed to implement the takeover request transition, from the ADAS control system to the driver, as required in the automated vehicles.

[0116] Finally, it is clear that modifications and variations may be made to the object of the present patent application described and illustrated herein without departing from the protective scope of the present invention as defined in the appended claims.

Claims

CLAIMS1. A computer program product (7) comprising instructions which, when executed by electronic processing resources (10) of a system (1) for online scoring sleep stages among Awake (W), Light Sleep (LS), Deep Sleep (DS) and REM sleep (R) of a subject, cause the electronic processing resources (10) to:- receive (S10-S14) at least one physiological signal of a subject;- process (S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject;- output (S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject; and- determine (S20, S23, S25, S28, S29) the status of the subject based on the outputted score, wherein the computer program product (7) is further configured to, when executed, cause the electronic processing resources (10) to:- determine a quantity related to the time needed for a user to resume control of a vehicle (TOT) after being asleep; and- provide assistance to the subject based on the quantity related to the time needed for a user to resume control of a vehicle after being asleep (TOT).

2. The computer program product (7) according to claim 1, wherein the at least one physiological signal is indicative of vital signs.

3. The computer program product (7) according to claim 2, wherein the vital signs comprise heart rate (HR) and / or heart rate variability (HRV) signals.

4. The computer program product (7) according to any one of the previous claims and further configured to, when executed, to cause the electronic processing resources (10) to determine (S14) if the subject is awake (W) at a first instant (i-1).

5. The computer program product (7) according to claim 4, wherein, in order to process (S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classifyit into one of different classes associated with the sleep stages of a subject, the software (7) is configured to cause the electronic processing resources (10) to:- if the subject is determined to be awake (W) in the first instant (i-1), determine (S15-S18) if one or more values of the vital signs is monotonically reducing for a given amount of observations (N); and- if the one or more values of the vital signs are determined to be reducing, determine (S 19) whether in a time frame the one or more vital signs satisfy a stabilization requirement, and wherein in order to output (S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software (7) is configured to cause the electronic processing resources (10) to, if the one or more vital signs satisfy the stabilization requirement in the time frame, determine (S20) that the subject is in a light sleep state (LS).

6. The computer program product (7) according to claim 5, wherein, in order to process (S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software (7) is configured to cause the electronic processing resources (10) to determine (S21-S22) if the values of the one or more vital signs are above a respective thresholds (HRTh, HRVTh), and wherein, in order to output (S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software (7) is configured to cause the electronic processing resources (10) to, if the values of the one or more vital signs are determined to be above the respective thresholds, determine (S23) that the subject in the awake state (W).

7. The computer program product (7) according to claim 6, wherein if the values of the one or more vital signs are above a respective thresholds (HRTh, HRVTh), in order to process (S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software(7) is configured to cause the electronic processing resources (10) to:- determine (S24) one or more statistical quantities related to a further quantity related to the vital signs (HRSmooth); and- determine (S24) if the values of the one or more statistical quantities related to the further quantity related to the vital signs (HRSmooth) are below respective threshold values (HRSmoothTh), and wherein, in order to output (S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software (7) is configured to cause the electronic processing resources (10) to, if it is determined that the values of the one or more statistical quantities related to the further quantity related to the vital signs (HRSmooth) are below respective threshold values (HRSmoothTh), determine (S25) that the subject is in a deep sleep state (DS).

8. The computer program product (7) according to claim 7, wherein, in order to process (S15-S19, S21-S22, S24, S26-S27) the received physiological signal to classify it into one of different classes associated with the sleep stages of a subject, the software (7) is configured to cause the electronic processing resources (10) to determine (S26-S27) if the values of the one or more statistical quantities are comprised between a first and a second threshold (SDNNThl, SDNNTh2), and wherein, in order to output (S20, S23, S25, S28, S29) a scoring of the determined sleep state of the subject and determine (S20, S23, S25, S28, S29) the status of the subject based on the outputted score, the software (7) is configured to cause the electronic processing resources (10) to:- if the values of the one or more statistical quantities are comprised between the first and the second threshold (SDNNThl, SDNNTh2), determine (S28) that the subject is in the REM state (R); and- if the values of the one or more statistical quantities are not comprised between the first and the second threshold (SDNNThl, SDNNTh2), determine (S29) that the subject is in the light sleep state (LS).

9. The computer program product according to any one of the preceding claims,wherein, in order to determine a quantity related to the time needed for a user to resume control of a vehicle (TOT) after being asleep, the software (7) is configured to cause the electronic processing resources (10) to determine a takeover transition time (TOT) of the subject.

10. System (1) for online scoring sleep stages among Awake (W), Light Sleep (LS), Deep Sleep (DS) and REM sleep (R) of a subject comprising electronic processing resources (10) configured to store and execute a computer program product according to any one of the preceding claims.

Images