Systems and methods for crash risk forecasting in traffic monitoring systems

Abstract

A crash-risk monitoring and mitigation system includes a main model that analyzes historic traffic data over an evaluation period to determine a crash-risk per road segment per time-segment. The crash-risk indicates a likelihood of vehicle collisions over the evaluation period. A surrogate model analyzes the historic traffic data to generate forecast traffic data. The forecast traffic data characterizes predicted traffic per road segment per future time-segment over a forecast period. The main model determines a crash-risk forecast by analyzing the forecast traffic data as if the forecast traffic data were historic traffic data. The crash-risk forecast indicates the likelihood of vehicle collisions over the forecast period. One or more driver assistance systems and / or traffic control mechanisms are operated in accordance with the crash-risk forecast so as to mitigate the likelihood of vehicle collisions.

Description

Attorney Docket No.: 113607.PH738WOSYSTEMS AND METHODS FOR CRASH RISK FORECASTING IN TRAFFIC MONITORING SYSTEMSCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 695, 197, filed Speptember 16, 2024, the entire disclosure of which is expressly incorporated by reference herein.BACKGROUND

[0002] The present invention relates to traffic monitoring systems and methods, and more particularly to such systems and methods providing crash risk forecasting.

[0003] Road safety is a critical concern worldwide. Traditional methods of determining crash risks rely on historic data and static factors, often overlooking dynamic and temporal aspects that significantly influence crash probabilities. There is a need for systems that can dynamically analyze and predict crash risks with high accuracy.BRIEF SUMMARY OF THE INVENTION

[0004] Systems and methods are disclosed for a traffic monitoring system that quickly and accurately predicts crash-risks on road segments by incorporating a main model in the form of a classifier, and at least one time series forecasting surrogate model. The system utilizes roadside sensors and / or connected vehicle data, environmental conditions, and temporal factors to provide real-time crash risk assessments. The integration of machine learning techniques with time series forecasting significantly improves prediction accuracy and reliability.

[0005] Other objects, advantages and novel features of the present invention will become apparent from the following detailed description of one or more preferred embodiments when considered in conjunction with the accompanying drawings. It should be recognized that the one or more examples in the disclosure are non-limiting examples and that the present invention is intended to encompass variations and equivalents of these examples.Attorney Docket No.: 113607.PH738WOBRIEF DESCRIPTION OF THE DRAWINGS

[0006] The features, objects, and advantages of the present invention will become more apparent from the detailed description, set forth below, when taken in conjunction with the drawings, in which like reference characters identify elements correspondingly throughout.

[0007] FIG. 1 is a schematic representation of a traffic monitoring system in accordance with one or more aspects of the invention.

[0008] FIG. 2 is a system architecture in accordance with one or more aspects of the invention;

[0009] FIG. 3 is an example table of historic traffic data for a sample road segment;

[0010] FIG. 4 is an example table of forecast traffic data for the sample road segment;

[0011] FIG. 5 is a schematic representation of exemplary interactions between exemplary modules of the system architecture in accordance with one or more aspects of the invention; and

[0012] FIG. 6 is an example table of crash-risk forecasts for the sample road segment.DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] The above described drawing figures illustrate the present invention in at least one embodiment, which is further defined in detail in the following description. Those having ordinary skill in the art may be able to make alterations and modifications to what is described herein without departing from its spirit and scope. While the present invention is susceptible of embodiment in many different forms, there is shown in the drawings and will herein be described in detail at least one preferred embodiment of the invention with the understanding that the present disclosure is to be considered as an exemplification of the principles of the present invention, and is not intended to limit the broad aspects of the present invention to any embodiment illustrated.

[0014] In accordance with the practices of persons skilled in the art, the invention is described below with reference to operations that are performed by a computer system or a like electronic system. Such operations are sometimes referred to as being computer-executed. It will be appreciated that operations that are symbolically represented include the manipulationAttorney Docket No.: 113607.PH738WO by a processor, such as a central processing unit, of electrical signals representing data bits and the maintenance of data bits at memory locations, such as in system memory, as well as other processing of signals. The memory locations where data bits are maintained are physical locations that have particular electrical, magnetic, optical, or organic properties corresponding to the data bits.

[0015] When implemented in software, code segments perform certain tasks described herein. The code segments can be stored in a processor readable medium. Examples of the processor readable mediums include an electronic circuit, a semiconductor memory device, a read-only memory (ROM), a flash memory or other non-volatile memory, a floppy diskette, a CD-ROM, an optical disk, a hard disk, etc.

[0016] In the following detailed description and corresponding figures, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it should be appreciated that the invention may be practiced without such specific details. Additionally, well-known methods, procedures, components, and circuits have not been described in detail.

[0017] The present invention generally relates to traffic monitoring systems and methods, and more particularly to such systems and methods providing crash risk forecasting.

[0018] FIG. 1 is a schematic representation of a traffic monitoring system 100 in accordance with one or more aspects of the invention. The traffic monitoring system comprises one or more traffic sensors 120 communicatively coupled to a system server 140, via a network 180. The system server may also be communicatively coupled to one or more devices 160 via the network. The traffic monitoring system generally enables the collection of traffic related data for transmission, via the network, to the system server. The traffic monitoring system also generally enables user access to the traffic related data stored on the system server, via the coupled devices.

[0019] The traffic sensors may comprise a sensor device 122, a controller 124, a memory 126, and a transceiver 128, each communicatively coupled to a common data bus 129 that enables data communication between the respective components.

[0020] The sensor device may include one or more image sensors, acceleration sensors, position sensors, doppler sensors, radar sensors, RFID tag sensors, induction loop sensors, orAttorney Docket No.: 113607.PH738WO any other types of sensors configured to capture data reflecting traffic and / or traffic-related conditions. The traffic and / or traffic-related conditions may include but are not limited to: the occurrence of vehicle collisions, vehicle count / volume and speed, as well as road and / or weather conditions.

[0021] Accordingly, the traffic sensors are preferably positioned at various locations that facilitate monitoring of the traffic and / or traffic-related conditions. For example, the traffic sensors may be positioned at various roadway locations where the traffic is to be monitored, and / or may be positioned in / on one or more of the vehicles constituting the traffic to be monitored. The sensor devices are preferably positioned such that the captured data includes the speed and / or count / volume of the passing vehicles, as well as other road and / or weather conditions. The captured data is preferably timestamped so as to reflect a date / time at which the data was captured. The captured data is also preferably associated with a location corresponding to the location of the monitored traffic. The location may be a location of the traffic sensor generating the captured data and / or the location of the road segment on which the traffic occurs. It will be understood that an identifier for the traffic sensor and / or the road segment may be used as a proxy for the location — via which the corresponding location of the traffic sensor and / or road segment may be looked-up using a referential table, preferably stored at the system server.

[0022] The controller may include processing software for processing the data captured by the sensor device so as to generate traffic-related data characterizing the traffic and / or traffic- related conditions. In some embodiments, the data may be processed so as to characterize the traffic and / or traffic-related conditions per road segment and / or per time segment. The controller may be embodied, collectively or individually, as one or more processors programmed to carry out the functions of the unit in accordance software stored in the memory. Each processor may be a standard processor, such as a central processing unit (CPU), or a dedicated processor, such as an application-specific integrated circuit (ASIC) or field programable gate array (FPGA), or portion thereof.

[0023] The memory stores software and data that can be accessed by the processor(s), and includes both transient and persistent storage. The transient storage is configured to temporarily store data being processed or otherwise acted on by other components, and may include a data cache, RAM or other transient storage types. The persistent storage is configuredAttorney Docket No.: 113607.PH738WO to store software and data until deleted. The memory is accordingly configured to store the data and information described herein.

[0024] The transceiver communicatively couples the traffic sensor to the network so as to enable data transmission therewith. The network may be any type of network, wired or wireless, configured to facilitate the communication and transmission of data, instructions, etc., and may include a local area network (LAN) (e.g., Ethernet or other IEEE 802.03 LAN technologies), Wi-Fi (e.g., IEEE 802.11 standards, wide area network (WAN), virtual private network (VPN), global area network (GAN)), a cellular network, or any other type of network or combination thereof.

[0025] The system server may include one or more server computers connected to the network. Each server computer may include computer components, including one or more processors 142, memories 144, displays 146, input / output interfaces 148, transceivers 149, and may also include software instructions and data for executing the functions of the server described herein. The servers may also include one or more storage devices 145 configured to store large quantities of data and / or information, and may further include one or more databases. For example, the storage device may be a collection of storage components, or a mixed collection of storage components, such as ROM, RAM, hard-drives, solid-state drives, removable drives, network storage, virtual memory, cache, registers, etc., configured so that the server computers may access it. The storage components may also support one or more databases for the storage of data therein.

[0026] The system server is generally configured to provide centralized support for the traffic sensors. The system server is configured to receive data from each of the traffic sensors, and to store the data for users to access via the devices. In some embodiments, the data received from the traffic sensors may be stored in the one or more databases.

[0027] The system server may include one or more crash-risk monitoring software applications, stored in the memory, which software applications, when executed by the processor configures the server computer to host and / or otherwise support a crash-risk monitoring platform. The crash-risk monitoring platform may be an online platform (e.g., a website) or a local platform (e.g., a closed computer network).

[0028] The crash-risk monitoring platform may be generally configured to analyze the traffic -related data so as to generate one or more crash-risk forecast reports, alerts and / orAttorney Docket No.: 113607.PH738WO signals that generally indicate the likelihood of one or more vehicle collisions in one or more road segments over a forecast period (e.g., the next hour, day, etc.). The crash-risk monitoring platform may further be configured to permit users, via the devices, to interact with data stored by the system server. In particular, the traffic monitoring platform may support a graphical user interface that permits users to receive crash-risk alerts and view crash-risk reports generated based on the one or more crash-risk forecasts. The crash-risk reports may identify crash-risks and / or crash-risk scenarios for road segments per future time-segment over the forecast period. This may include, for example identifying the predicted traffic conditions (e.g., in terms of average vehicle speed and / or count / volume) contributing to the predicted crash-risk and / or traffic conditions. Thus, for example, the crash-risk report may identify that the predicted average vehicle speed on a certain road segment in the next hour indicates a crashrisk for that road segement in the next hour. The alerts and / or signals may be generated based on the reports and may include information from the reports.

[0029] Moreover, in some embodiments, the reports, alerts and / or signals may be transmitted to one or more vehicles so as to cause the vehicles to adjust control of one or more vehicle systems. For example, automous driving systems may be adjusted so as to reduce vehicle speed on the road segment and thereby avoid the predicted traffic conditions correlated with the crash-risk report, alert and / or signal. As another example, the automated operation of traffic lights and other traffic control systems can be adjusted to reduce vehicle speed on the road segment to similar effect. As another example, vehicle navigation systems may be adjusted so as to reduce vehicle count / volume on the road segment. As will be apparent from the descriptions herein, a feedback loop may be effectuated with such automated systems for controlling traffic to avoid predicted traffic conditions, in which updated traffic-related data is used to update the crash -risk forecast reports, alerts and / or signals, such that the automated traffic control measures can be rolled-back when the crash-risk is no longer identified for the road segment.

[0030] The devices are generally computing devices, and may include mobile (e.g., laptop computer, tablet computer, smartphone, PDA, wearable, etc.) or stationary (e.g., desktop computer, etc.), multi-purpose or dedicated, devices configured to communicate data and information with the system server. Moreover, the devices may be on-vehicle devices (e.g., vehicle infotainment systems, vehicle navigation systems, driver assistance systems, etc.) or off-vehicle devices (e.g., traffic planning / control systems, emergency response systems, etc.).Attorney Docket No.: 113607.PH738WO

[0031] The devices may include components typically associated with such devices, such as one or more processors, physical memories, software instructions, data, displays, and interfaces. The devices may further include one or more software applications, stored in memory, which software applications, when executed by the processor, configures the devices to function as described herein. In particular, the devices are configured to allow the users to interact with the traffic monitoring platform, as described herein.

[0032] FIG. 2 is a system architecture 200 in accordance with one or more aspects of the invention. The system architecture generally enables the functions of the traffic sensors and / or the server system in accordance with the crash-risk monitoring platform. The system architecture comprises one or more hardware and / or software modules configured to implement the functions described herein. More particularly, the system architecture generally comprises a data collection module 210, a database 220, a data processing unit 230, a main model 240, a surrogate model 250, and an output interface 270.

[0033] The data collection module is generally configured to receive the traffic-related data from the traffic sensors and retrievably store the traffic-related data in the database for use in accordance with at least one embodiment, as described herein.

[0034] The data processing unit is generally configured to process the traffic-related data so as to generate historic traffic data that statistically characterizes the monitored traffic per road segment per time- segment. The time- segment may be any amount of time, but is preferably 1-hour. Thus, in at least one embodiment, the historic traffic data statistically characterizes the monitored traffic in 1-hour increments.

[0035] The historic traffic data may statistically characterize the monitored traffic according to one or more traffic parameters having corresponding values that reflect statistical traffic conditions derived from the underlying traffic-related data. For example, the traffic parameters of a given road segment may include, but are not limited to: an average vehicle speed of the road segment, and / or an average vehicle count / volume of the road segment. It will be understood that, while the exemplary traffic parameters are described in terms of averages, any statistical characterization may be used without departing from the scope of the invention. Moreover, the statistical characterization may include one or more indicators of statistical variance (e.g., standard deviation, z-score, etc.).Attorney Docket No.: 113607.PH738WO

[0036] In some embodiments, the historic traffic data may also include one or more roadcondition parameters having corresponding values that reflect statistical road conditions derived from the underlying traffic-related data. For example, the road-condition parameters may include, but are not limited to: an average amount of ice, number of potholes, etc., on the road segment; an average composition of the road segment; and / or an average grade, curvature, etc., of the road segment. It will be understood that, while the exemplary road-condition parameters are described in terms of averages, any statistical characterization may be used without departing from the scope of the invention. Moreover, the statistical characterization may include one or more indicators of statistical variance (e.g., standard deviation, z-score, etc.).

[0037] In some embodiments, the historic traffic data may also include one or more weather-condition parameters having corresponding values that reflect statistical weather conditions derived from the underlying traffic-related data. For example, the weathercondition parameters may include, but are not limited to: an average precipitation (e.g., rainfall, snowfall, etc.) on the road segment; an average visibility on the road segment; an average wind speed / direction on the road segment; and / or an average temperature on the road segment. It will be understood that, while the exemplary weather-condition parameters are described in terms of averages, any statistical characterization may be used without departing from the scope of the invention. Moreover, the statistical characterization may include one or more indicators of statistical variance (e.g., standard deviation, z-score, etc.).

[0038] In some embodiments, the historic traffic data may also include one or more temporal parameters having corresponding values that reflect the temporal conditions derived from the underlying traffic-data. For example, the temporal parameters may include, but are not limited to: whether the time-segment(s) of the historic traffic data fall on a weekday / weekend / holiday; the day of the week that the time-segment(s) fall on; the part of the day (e.g., morning, afternoon, evening, night, etc.) that the time-segment(s) fall on; and / or the specific time-segment(s) (e.g., between 6:00 AM and 7:00 AM) of the historic traffic data.

[0039] FIG. 3 illustrates a simplified example table 300 of historic traffic data for exemplary road segment Rt. As can be seen, the historic traffic data includes the exemplary traffic parameters of average speed V and average vehicle count C. Each of the traffic parameters of average speed V and average vehicle count C have corresponding values (i.e.,to Cn, respectively) for time-segments 7 to Tn, that statistically characterize theAttorney Docket No.: 113607.PH738WO monitored traffic according to those traffic parameters, which values reflect the statistical traffic conditions derived from the underlying traffic -related data. As noted herein, the table may include additional and / or other parameters and values without departing from the scope of the invention.

[0040] The processing of the traffic-related data is preferably carried out by the data processing unit in real-time or in near real-time with the capture and / or transmission of the traffic -related data. In some embodiments, the processing may be substantially contemporaneous with the end of the corresponding time-segment. It will be understood, however, that the traffic-related data is preferably processed such that there is minimal delay between the end of the corresponding time-segment and the processing of the traffic -related data. For example, the traffic-related data for the 6:00 AM to 7:00 AM time-segment should be processed (and the historic traffic data should be generated) at or soon after 7:00 AM. Longer delays in processing are, however, expressly contemplated. The historic traffic data may be retrievably stored in the database for use in accordance with at least one embodiment, as described herein.

[0041] The main model is generally configured to analyze the historic traffic data over an evaluation period so as to determine a crash-risk per road segment per time-segment. The crash-risk indicates a likelihood of one or more vehicle collisions over the evaluation period. The crash-risk determination may be a binary determination (e.g., crash-risk vs. no crash-risk) or may be a stepped determination (e.g., low / medium / high, percentage, etc.) of the likelihood of vehicle collision. Thus, the main model is generally used, for example, to identify unreported collisions and / or high crash-risk trends for remedial action.

[0042] In operation, the main model may be configured to identify one or more crash-risk scenarios within the historic traffic data. The crash-risk scenarios may be values or combinations of values for one or more of the traffic parameters that are historically indicative of vehicle collisions. For example, crash-risk scenarios may include, but are not limited to: the average vehicle count exceeding a predetermined count threshold, and / or the average vehicle speed exceeding a predetermined speed threshold.

[0043] The determination of the crash-risk for the road segment may be based on the crashrisk scenarios identified for the road segment and / or or the lack thereof. Thus, for example, where one or more crash-risk scenarios are identified for the road segment, the crash-risk mayAttorney Docket No.: 113607.PH738WO be determined. In some embodiments, the crash-risk may be determined based on the number of crash-risk scenarios identified and / or the severity of the crash-risk scenarios identified. Thus, the sensitivity of the determination is possible by adjusting one or more thresholds for the number and / or severity of crash-risk scenarios identified.

[0044] The main model may comprise an artificial intelligence trained on the historic traffic data and / or the traffic-related data to identify the crash-risk scenarios in the historic traffic data. The details of training the artificial intelligence is not the subject of this disclosure.

[0045] In some embodiments, the crash-risk scenarios may also be defined by values or combinations of values for one or more of the road-condition, weather-condition, and / or temporal parameters that are historically indicative of vehicle collisions. Thus, for example, crash risk scenarios may include, but are not limited to: the average vehicle count exceeding the predetermined count threshold, and / or the average vehicle speed exceeding the predetermined speed threshold, under certain road-conditions, weather-conditions and / or temporal conditions. In this manner, crash-risk scenarios may be more granularly defined. Moreover, in this manner, the cause(s) of the crash-risk may be identified. For example, where the average vehicle count exceeds the count threshold, it can be determined that the count / volume of vheicles is likely the cause of the identified crash-risk. Accordingly, as discussed herein, steps may be taken to mitigate the risk.

[0046] The evaluation period may be any period over which the crash-risk determination is desired, but is preferably a multiple of the unit time. It will be understood, however, that the evaluation period is necessarily historic, because it corresponds to a period for which the historic traffic data exists. As such, in determining the crash-risk for the evaluation period, the main model “looks backward” at the evaluation period — and is generally not configured to “look forward” via analyzing the historic traffic data over the evaluation period.

[0047] Accordingly, the system architecture relies on the surrogate model to enable the main model to “look forward” without needing to fundamentally reconfigure the main model. The surrogate model is generally configured to analyze historic traffic data so as to generate forecast traffic data that statistically characterizes predicted traffic per road segment per future time-segment over a forecast period. As explained herein, this forecasting data can then be used by the main model to predict determine a crash-risk forecast for each of the road segments, which crash-risk forecast effectively “looks ahead” to predict a likelihood of one or moreAttorney Docket No.: 113607.PH738WO vehicle collisions over the forecast period. In general, the surrogate model may be any timeseries forecasting model that predicts at least one traffic parameter of the road segment per future time- segment over the forecast period.

[0048] The future time-segments are time-segments occurring after the most current timesegment for which historic traffic data is available. For example, where the historic traffic data is available only up to 7:00 AM (e.g., because the traffic-related data after 7:00 AM has not yet been captured), the future-time segments may be the 1-hour time-segments after 7:00 AM. Thus, the forecast period generally comprises some range of future time- segments occurring after the most current time-segment for which historic traffic data is available.

[0049] It will be understood that the forecast period may be any range of future timesegments for which the crash-risk forecasting is desired. Moreover, the forecast period may be according to a predetermined schedule, e.g., hourly, bi-hourly, daily, etc., such that the forecast traffic data is generated in accordance with the predetermined schedule — and / or vice versa. For example, where the forecast period is 1-hour long, the forecast traffic data for the 1-hour forecast period can be generated at the top of the hour (or just before). Similarly, where the forecast traffic data is set to be generated at the top of every hour (or just before), the forecast period can be set at 1-hour long. In this way, successively generated forecast traffic data can continue to outpace the historic traffic data being captured by the traffic sensors. It will be understood, however, that the forecast period need not directly correspond to the rate of forecast traffic data generation. For example, in some embodiments, the forecast traffic data may be generated hourly where the forecast period is several hours long. In this way, the forecast traffic data for future time-segments (e.g., several hours away) can be regularly updated in accordance with more recently captured historic traffic data (e.g., of the most recent hour). It will be further understood that other schedules, including random schedules, may be used without departing from the scope of the invention.

[0050] The forecast traffic data may statistically characterize the predicted traffic according to the one or more traffic parameters via corresponding forecasted values that reflect predicted traffic conditions derived from the historic traffic data. For example, the forecast traffic data may predict, but is not limited to predicting: the average vehicle speed of the road segment, and / or the average vehicle count / volume of the road segment, per future timesegment over the forecast period. It will be understood that, while the exemplary traffic parameters are described in terms of averages, any statistical characterization may be usedAttorney Docket No.: 113607.PH738WO without departing from the scope of the invention. Moreover, the statistical characterization may include one or more indicators of statistical variance (e.g., standard deviation, z-score, etc.).

[0051] In some embodiments, the forecast traffic data may also include the one or more road-condition parameters having corresponding forecasted values that reflect predicted roadconditions derived from the historic traffic data. For example, the forecast traffic data may predict, but is not limited to predicting: the average amount of ice, number of potholes, etc., on the road segment; the average composition of the road segment; and / or the average grade, curvature, etc., of the road segment. It will be understood that, while the exemplary roadcondition parameters are described in terms of averages, any statistical characterization may be used without departing from the scope of the invention. Moreover, the statistical characterization may include one or more indicators of statistical variance (e.g., standard deviation, z-score, etc.).

[0052] In some embodiments, the forecast traffic data may also include the one or more weather-condition parameters having corresponding forecasted values that reflect predicted weather-conditions derived from the historic traffic data. For example, the forecast traffic data may predict, but is not limited to predicting: the average precipitation (e.g., rainfall, snowfall, etc.) on the road segment; the average visibility on the road segment; the average wind speed / direction on the road segment; and / or the average temperature on the road segment. It will be understood that, while the exemplary weather-condition parameters are described in terms of averages, any statistical characterization may be used without departing from the scope of the invention. Moreover, the statistical characterization may include one or more indicators of statistical variance (e.g., standard deviation, z-score, etc.).

[0053] In some embodiments, the forecast traffic data may also include the one or more temporal parameters having corresponding forecasted values that reflect temporal conditions of the forecast traffic data. For example, the forecast traffic data may indicate, but is not limited to: whether the future time-segment(s) fall on a weekday / weekend / holiday; the day of the week that the future time-segment(s) fall on; the part of the day (e.g., morning, afternoon, evening, night, etc.) that the future time-segment(s) fall on; and / or the specific future time-segment(s) of the forecasted traffic data.Attorney Docket No.: 113607.PH738WO

[0054] FIG. 4 illustrates a simplified example table 400 of forecast traffic data for exemplary road segment R^ . As can be seen, the forecast traffic data includes the exemplary traffic parameters of average speed V and average vehicle count C as the corresponding historic traffic data from which the forecast traffic data is derived. However, with the forecast traffic data, each of the traffic parameters now have corresponding forecasted values (i.e.,and C to Cn' , respectively) for future time-segments T1' to T^, that statistically characterize the predicted traffic according to those traffic parameters, which forecasted values reflect the predicted traffic conditions derived from the historic traffic data. As noted herein, the table may include additional and / or other parameters and values without departing from the scope of the invention. The forecast traffic data may be retrievably stored in the database for use in accordance with at least one embodiment, as described herein.

[0055] Turning now to FIG. 5, the main model is further generally configured to analyze the forecasted traffic data over the forecast period so as to determine the crash-risk forecast per road segment per future time-segment. As previously noted, the crash-risk forecast indicates the likelihood of one or more vehicle collisions over the forecast period, and may generally correspond to the identification of the crash-risk for future time-segments. In this manner, the main model need not be fundamentally modified to predict the risk of vehicle collision. In other words, the main model is supplied with the forecasted traffic data as if it were historic traffic data, and from the forecasted traffic data, the main model generates the crash-risk forecast as if it were identifying the crash-risk over the evaluation period.

[0056] Accordingly, the main model may be configured to identify one or more of the crash-risk scenarios within the forecast traffic data in a substantively identical manner as the main model identifies crash-risk scenarios within the historic traffic data. In particular, as applied to the forecasted traffic data, the crash-risk scenarios identified for the forecast period may correspond to forecasted values or combinations of forecasted values for one or more of the traffic parameters that would be similarly indicative of vehicle collisions in the historic traffic data. For example, as discussed herein, the crash-risk scenarios may include, but are not limited to: the average vehicle count exceeding the predetermined count threshold, and / or the average vehicle speed exceeding the predetermined speed threshold. Thus, if the forecasted values exceed such thresholds, the corresponding crash-risk scenario will be identified similarly to such identification in the case of historic values.Attorney Docket No.: 113607.PH738WO

[0057] Similar to the identification of historic crash-risk scenarios, in some embodiments, the crash-risk scenarios may also be defined by values or combinations of values for one or more of the road-condition, weather-condition, and / or temporal parameters. Thus, for example, crash-risk scenarios within the forecast traffic data may likewise include, but are not limited to: the average vehicle count exceeding the predetermined count threshold, and / or the average vehicle speed exceeding the predetermined speed threshold, under certain roadconditions, weather-conditions and / or temporal conditions. In this manner, crash-risk scenarios may be more granularly defined.

[0058] FIG. 6 illustrates an exemplary table 600 with determined crash-risk forecasts for exemplary road segment Rt. As can be seen, the crash-risk forecast comprises a crash-risk existence indicator 610 and / or a crash-risk probability indicator 620 (both are shown in FIG. 6 for simplicity sake). The crash-risk existence indicator may be a binary indicator that indicates whether the main model has predicted that one or more vehicle collisions are likely during the future time-segment. The crash-risk probability indicator may be a stepped indicator (e.g., low / medium / high, percentage, etc.) that indicates the likelihood of one or more vehicle collisions during the future time-segment, as predicted by the main model.

[0059] Turning back to FIGS. 2 and 5, the output interface is generally configured to generate the one or more crash-risk forecast reports, alerts and / or signals based on the crashrisk forecasts determined by the main model. The crash-risk forecast reports, alerts and / or signal may indicate the crash-risk forecast as determined for one or more road segments over the forecast period.

[0060] In general, the crash-risk forecast reports, alerts and / or signals may be transmitted the one or more devices so as to trigger responsive action by the devices. Such responsive action may include, but is not limited to: displaying or otherwise communicating the crash-risk report, alert and / or signal to a user of the device; adjusting the operation of a driver assistance system (e.g., navigation system, autonomous driving system, etc.) of one or more vehicles based on the report, alert and / or signal; and / or adjusting the operation of one or more traffic control mechanisms (e.g., traffic lights, etc.) based on the report, alert and / or signal.

[0061] The embodiments described in detail above are considered novel over the prior art and are considered critical to the operation of at least one aspect of the described systems, methods and / or apparatuses, and to the achievement of the above described objectives. TheAttorney Docket No.: 113607.PH738WO words used in this specification to describe the instant embodiments are to be understood not only in the sense of their commonly defined meanings, but to include by special definition in this specification: structure, material or acts beyond the scope of the commonly defined meanings. Thus, if an element can be understood in the context of this specification as including more than one meaning, then its use must be understood as being generic to all possible meanings supported by the specification and by the word or words describing the element.

[0062] The definitions of the words or drawing elements described herein are meant to include not only the combination of elements which are literally set forth, but all equivalent structure, material or acts for performing substantially the same function in substantially the same way to obtain substantially the same result. In this sense, it is therefore contemplated that an equivalent substitution of two or more elements may be made for any one of the elements described and its various embodiments or that a single element may be substituted for two or more elements.

[0063] Changes from the subject matter as viewed by a person with ordinary skill in the art, now known or later devised, are expressly contemplated as being equivalents within the scope intended and its various embodiments. Therefore, obvious substitutions now or later known to one with ordinary skill in the art are defined to be within the scope of the defined elements. This disclosure is thus meant to be understood to include what is specifically illustrated and described above, what is conceptually equivalent, what can be obviously substituted, and also what incorporates the essential ideas.

[0064] Furthermore, the functionalities described herein may be implemented via hardware, software, firmware or any combination thereof, unless expressly indicated otherwise. If implemented in software, the functionalities may be stored in a memory as one or more instructions on a computer readable medium, including any available media accessible by a computer that can be used to store desired program code in the form of instructions, data structures or the like. Thus, certain aspects may comprise a computer program product for performing the operations presented herein, such computer program product comprising a computer readable medium having instructions stored thereon, the instructions being executable by one or more processors to perform the operations described herein. It will be appreciated that software or instructions may also be transmitted over a transmission medium as is known in the art. Further, modules and / or other appropriate means for performing theAttorney Docket No.: 113607.PH738WO operations described herein may be utilized in implementing the functionalities described herein.

[0065] The foregoing disclosure has been set forth merely to illustrate the invention and is not intended to be limiting. Since modifications of the disclosed embodiments incorporating the spirit and substance of the invention may occur to persons skilled in the art, the invention should be construed to include everything within the scope of the described embodiments and equivalents thereof.

Claims

Attorney Docket No.: 113607.PH738WOCLAIMS1. A crash-risk monitoring and mitigation system, comprising: a computer system, including: a main model that retrieves and analyzes historic traffic data over an evaluation period to determine a crash-risk per road segment per time-segment, wherein the crashrisk indicates a likelihood of one or more vehicle collisions over the evaluation period, and a surrogate model that retrieves and analyzes the historic traffic data to generate forecast traffic data and provide the forecast traffic data to the main model, wherein the forecast traffic data characterizes predicted traffic per road segment per future timesegment over a forecast period, wherein the main model is configured to determine a crash-risk forecast by analyzing the forecast traffic data as if the forecast traffic data were historic traffic data, the crash-risk forecast indicating the likelihood of one or more vehicle collisions over the forecast period; and one or more driver assistance systems and / or traffic control mechanisms configured to be operated in accordance with the crash-risk forecast so as to mitigate the likelihood of the one or more vehicle collisions.

2. The crash-risk monitoring and mitigation system of claim 1, wherein the crash-risk forecast indicates a binary likelihood of one or more vehicle collisions.

3. The crash-risk monitoring and mitigation system of claim 1, wherein the crash-risk forecast indicates a stepped likelihood of one or more vehicle collisions.

4. The crash-risk monitoring and mitigation system of claim 1, wherein the crash-risk forecast is determined based on an identification of one or more crash risk scenarios per road segment per future time- segment.

5. The crash-risk monitoring and mitigation system of claim 4, wherein the crash-risk forecast is determined based on the severity of identified crash risk scenarios.Attorney Docket No.: 113607.PH738WO6. The crash-risk monitoring and mitigation system of claim 4, wherein the crash-risk forecast is determined based on the number of identified crash risk scenarios.

7. The crash-risk monitoring and mitigation system of claim 1, wherein the surrogate model is a time-series forecasting model.

8. The crash-risk monitoring and mitigation system of claim 1, wherein the forecast traffic data is generated according to a predetermined schedule.

9. The crash-risk monitoring and mitigation system of claim 1 , wherein the forecast traffic data outpaces the historic traffic data.

10. A crash-risk monitoring and mitigation method, comprising: providing a main model that retrieves and analyzes historic traffic data over an evaluation period to determine a crash-risk per road segment per time-segment, wherein the crash-risk indicates a likelihood of one or more vehicle collisions over the evaluation period; providing a surrogate model that retrieves and analyzes the historic traffic data to generate forecast traffic data and provide the forecast traffic data to the main model, wherein the forecast traffic data characterizes predicted traffic per road segment per future time- segment over a forecast period; determining, by the main model, a crash-risk forecast by analyzing the forecast traffic data as if the forecast traffic data were historic traffic data, the crash-risk forecast indicating the likelihood of one or more vehicle collisions over the forecast period; and operating one or more driver assistance systems and / or traffic control mechanisms in accordance with the crash-risk forecast so as to mitigate the likelihood of the one or more vehicle collisions.

11. The crash-risk monitoring and mitigation method of claim 10, wherein the crash-risk forecast indicates a binary likelihood of one or more vehicle collisions.

12. The crash-risk monitoring and mitigation method of claim 10, wherein the crash-risk forecast indicates a stepped likelihood of one or more vehicle collisions.Attorney Docket No.: 113607.PH738WO13. The crash-risk monitoring and mitigation method of claim 10, wherein the crash-risk forecast is determined based on an identification of one or more crash risk scenarios per road segment per future time- segment.

14. The crash-risk monitoring and mitigation method of claim 13, wherein the crash-risk forecast is determined based on the severity of identified crash risk scenarios.

15. The crash-risk monitoring and mitigation method of claim 13, wherein the crash-risk forecast is determined based on the number of identified crash risk scenarios.

16. The crash-risk monitoring and mitigation method of claim 10, wherein the surrogate model is a time-series forecasting model.

17. The crash-risk monitoring and mitigation method of claim 10, wherein the forecast traffic data is generated according to a predetermined schedule.

18. The crash-risk monitoring and mitigation method of claim 10, wherein the forecast traffic data outpaces the historic traffic data.

Images