Method of and system for detecting vehicles

Abstract

A method and system of detecting vehicles including receiving, from a first sensor device at a first location, a first noise level and a first speed of ones of a plurality of vehicles traveling past the first location, receiving, from a second sensor device at a second location, a second noise level and a second speed of ones of the plurality of vehicles traveling past the second location, and classifying, by a processor, one or more of the plurality of vehicles as being either an electric vehicle or an internal-combustion-engine vehicle based on a difference between the first noise level and the second noise level of the vehicle, and further based on the first speed, the second speed, or both the first and second speed of the vehicle.

Description

[0001] METHOD OF AND SYSTEM FOR DETECTING VEHICLES

[0002] FIELD OF THE INVENTION

[0003] The present disclosure is generally directed to systems and methods for detecting vehicles using, for example, radar and audio sensors. More specifically, the disclosure is directed to systems and methods for deteriming a number of electric vehicles (EVs) and internal-combustion engine vehicels (ICEVs).

[0004] BACKGROUND OF THE INVENTION

[0005] According to the National Highway Traffic Safety Administration (NHTSA), electric vehicles (HVs) and a combination of an EV and an internal-combustion-engine) (ICE) (commonly referred to as a hybrid vehicle (HV) cause twice as many vehicle-to- pedestrian accidents as ICE vehicles (ICEVs) when reversing or parking at slow speeds. In addition, an increase in EV / HV-to-bicycle accidents has been reported. As the number of EVs and HVs increase, there is a need to alert road users of the presence of HVs and EVs. Intelligent transportation systems (ITS) may also need to detect and track vehicle type. Furthermore, environmental impact, prioritized parking, driving, and traffic congestion mitigation may require monitoring of vehicle types on the road.

[0006] As EVs (including electric bicycles) increase and charging networks expand, forecasting charging demand may be required by energy providers to better optimize power distribution and infrastructure expansion (e.g., three EVs may experience a longer green light period that four ICEVs). Operators of EV Charging Stations (EVCS) may also require forecasting charging demand as they seek to find optimal locations for charging infrastructure. Despite an interest in modelling EV charger demand, data on private EV vehicle location is usually not publicly available which significantly limits the development of forecasting methods.

[0007] It is difficult to differentiate an HV from an ICEV from their images without license plate recognition. Thus, there is a need for a reliable and privacy-respecting way to determine an EV from an ICEV. SUMMARY OF THE INVENTION

[0008] In accordance with the concepts described herein, exemplary methods and systems determine a number of EVs and ICEVs on the road.

[0009] In accordance with the concepts described herein, exemplary methods and systems employ a noise sensor and a speed sensor to determine a number of EVs and ICEVs on the road.

[0010] In accordance with the concepts described herein, exemplary methods and systems count EVs and ICEVs to estimate a spatial -temporal traffic density of EVs and ICEVs in a privacy-sensitive manner.

[0011] BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The manner and process of making and using the disclosed embodiments may be appreciated by reference to the figures of the accompanying drawings. It should be appreciated that the components and structures illustrated in the figures are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the concepts described herein. Like reference numerals designate corresponding parts throughout the different views. Furthermore, embodiments are illustrated by way of example and not limitation in the figures, in which:

[0013] FIG. 1(a) is an illustration of an exemplary Outdoor Sensor Bundle (OSB) according to the present disclosure;

[0014] FIG. 1(b) is an exemplary chart of vehicle speed versus equivalent noise level for an EV / HV and two gasoline engines (GEs) according to the present disclosure;

[0015] FIG. 2 is an exemplary flowchart of a method of determining an EV according to the present disclosure;

[0016] FIG. 3 is an exemplary method of determining an EV and an ICEV according to the present disclosure;

[0017] FIG. 4 is an exemplary system for determining an EV and an ICEV according to the present disclosure; and

[0018] FIG. 5 is a block diagram of a processing device on which methods and processes disclosed herein can be implemented, according to some embodiments of the disclosure. DETAILED DESCRIPTION

[0019] FIG. 1(a) is an illustration of an exemplary first sensor device (e.g., an OSB) 101 according to the present disclosure. The first sensor device 101 includes a vibration sensor, radar, a tilt sensor, a microphone (e.g., an audio sensor) and a temperature sensor. The first sensor device 101 is shown with a streetlight pole 103 (or, in some cases, a roadlight pole or in proximity to a streetlight pole or a roadside pole). However, the present disclosure is not limited thereto.

[0020] An audio sensor mounted on a streetlight pole may detect a difference of sequential audio noise between an EV and an ICEV whenever vehicle speed is low (e.g., < lOkm / h) as monitored by the radar sensor in the first sensor device 101.

[0021] FIG. 1(b) is an exemplary chart of vehicle speed versus equivalent noise level for an EV / HV and two gasoline engines (GEs) according to the present disclosure. FIG. 1(b) shows the audio noise difference between two ICEVs GE1 and GE2 and an HV in EV mode.

[0022] In an embodiment, a system employs a radar, a first audio sensor, and a second audio sensor to detect and count EVs. The radar may detect vehicle speed. The first audio sensor and the second audio sensor may detect noise levels from EVs and ICEVs at various speeds. The first audio sensor and the second audio sensor may be mounted on a section of a road. The first audio sensor may be mounted at a streetlight closest to a road intersection. However, the present invention is not limited thereto. The second audio sensor may be mounted 50m away from the first audio sensor, and leading up to, the road intersection. Time-series audio noise data recorded by the first audio sensor and the second audio sensor along with the vehicle speed recorded by the radar enables classification (e.g., by a maximum likelihood (ML) method) to classify whether a vehicle is an EV or an ICEV, even if some of the EVs emit additive warning sounds. If there are multiple vehicles in the road, warning sounds may be filtered out as interference since a warning sound may impact the detection of an EV that does not emit a warning sound.

[0023] In embodiments, a noise differemce between an EV and an ICEV changes based in part on a speed of a respective vehicle. For example, an audio noise difference between an EV and an ICEV may be lOdB or greater whenever vehicle speed is below lOkm / h. An additive warning sound of an EV is significantly different from a sound made by an ICEV. Audio noise of an ICEV changes with ICEV speed and depends on an average wind speed, an amplitude modulation of the audio noise, ICEV velocity, minimum ambient sound level, and the ICEV type. An EV and an ICEV type may be an automobile, a truck, a bus, a motorcycle, or a bicycle. An ICEV may have a gasoline-powered engine or a hydrogen-powered engine. The present disclosure is capable of distinguishing between an ICEV powered by a gasoline engine and an ICEV powered by a hydrogen-powered engine. A hydrogen-powered ICEV may be preferred over a gasoline-powered ICEV because a hydrogen-powered ICEV pollutes less than a gasoline-powered ICEV.

[0024] Some hydrogen-powered cars such as TOYOTA MIRAI use a fuel cell to motivate an electrical motor in an EV. This first type of hydrogen-fuel-cell cars are hence quite silent and produce a soft motor hum which rises with the number or revolutions per minute (RPM). Other hydrogen powered cars such as the YARIS H2 feature a three cylinder turbocharged engine that bums hydrogen instead of gasoline and has an exhaust note somewhat similar to the sound of an ICEV. It is known that a hydrogen powered ICEV produces higher noise levels than a gasoline powered engine (e.g., Volkswagen Polo 1.4 may have 1.1 to 1.7 dBA mainly due to intensification of a 500 MHz band at speeds between 3000 and 4000 RPM. The highest noise difference between a gasoline-powered vehicle and a hydrogen-powered vehicle is at idling condition. Besides loudness, sharpness and roughness of the sound may also be used to distinguish between a gasoline-powered vehicle and a hydrogen-powered vehicle. The noise difference is most pronounced at low RPMs (e.g., 3000 RPM). Embodiments of the present disclosure may be used to detect different types of hydrogen-power cars.

[0025] FIG. 2 is an exemplary flowchart of a method of determining an EV according to the present disclosure. For example, the method may be applied to a scenario where there is a stoplight 201 at an intersection, a first set of sensors 203 attached to a first light pole 205 (e.g., Luminaire Li) in the vicinity of the stoplight 201, a second set of sensors 207 attached to a second light pole 209 (e.g., Luminaire L2) a distance away from, and oriented toward, the first light pole 205, and where vehicles 211, 213 approach the stoplight 201. However, the present disclosure is not limited thereto. The method may determine the type of vehicle (e.g., an EV or an ICEV) in any of three situations (e.g., a vehicle approaching the stoplight 201, a vehicle decelerating at the stoplight 201 when the stoplight 201 turns red, and a vehicle accelerating past the stoplight 201 when the stoplight 201 turns green).

[0026] For the situation where vehicles approach the stoplight 201, step 215 of the method comprises speed sensors (e.g., radar sensors) in each of the first set of sensors 203 and the second set of sensors 207 detecting a speed of each vehicle that passes the first set of sensors 203 and the second set of sensors 207, respectively. Step 217 of the method comprises audio sensors (e.g., microphones) in each of the first set of sensors 203 and the second set of sensors 207 recording noise of each vehicle that passes the first set of sensors 203 and the second set of sensors 207 for which a speed was detected, respectively. Step 219 of the method comprises comparing the noises recorded by the audio sensors (e.g., noise coming from fronts of the vehicles and noises coming from backs of the vehicles, respectively). Step 221 of the method comprises determining if the vehicles are EVs, respectively.

[0027] For the situation where vehicles decelerate at the stoplight 201 when the stoplight 201 turns red, step 223 of the method comprises speed sensors (e.g., radar sensors) in each of the first set of sensors 203 and the second set of sensors 207 detecting a speed of each vehicle that passes the first set of sensors 203 and the second set of sensors 207, respectively. Step 225 of the method comprises determining if the stoplight 201 is red and the speed of each vehicle is less than 10 kilometers per hour (km / h). If so, step 227 of the method determines if the number of vehicles is less than a user-definable number N (e.g., 10). If so, step 229 of the method comprises using an audio maximum likelihood (ML) model on audio of the vehicles recorded by audio sensors in each of the first set of sensors 203 and the second set of sensors 207 that pass the first set of sensors 203 and the second set of sensors 207 for which a speed was detected to determine a vehicle type (e.g., automobile, truck, bus, etc.), respectively. Step 231 of the method comprises computing an expected audio level with changing speed. Step 221 of the method is then repeated for this situation (e.g., determining if the vehicles are EVs, respectively).

[0028] For the situation where vehicles accelerate past stoplight 201 when stoplight 201 turns green, step 233 of the method comprises determining if the stoplight 201 turned green and a vehicle accelerates. If so, step 235 of the method comprises using an audio ML model on audio of the vehicles recorded by audio sensors in each of the first set of sensors 203 and the second set of sensors 207 to determine engine noise of each vehicle. Step 237 of the method comprises computing an expected engine noise level for acceleration for each vehicle. Step 221 of the method is then repeated for this situation (e.g., determining if the vehicles are EVs, respectively).

[0029] In an alternate embodiment, the method may employ audio data and vehicle speed to determine spatial -temporal traffic density of EVs and electric bikes. In an alternate embodiment, an ML classifier may analyze audio noise and compare a radar cross section obtained by sensors sensing vehicles from the front and from the back. As EVs have a front trunk (e.g., a frunk) instead of an engine in the front, EVs have less difference in their radar cross section from the front and from the back than ICEVs. Based upon this difference, an EV may be distinguished from an ICEV. The radar cross-section classification method also works when a vehicle is driven fast (e.g., where an audio-based classification of EV vs ICEV may not be possible).

[0030] Acceleration of a vehicle is a feature to distinguish between EVs and ICEVs. For instance, there may be a larger delay in acceleration for an ICEV than an EV. Radar may also be used to detect an absence of gear changes and, thereby, classify a vehicle as an EV. Radar may be used to monitor a vehicle's micromovements. An ICEV may run less smoothly than an EV and, thus, generate micro vibrations. As EVs are much heavier than ICEVs, a vibration sensor may be used by a ML classifier.

[0031] While an EV’s headlights are always powered by battery, an ICEV engine (via its alternator) provides power to its headlights. Hence during night hours, a light sensor may be used to classify EVs which may have a smoother light output as compared to alternator- powered headlights of an ICEV.

[0032] A radar sensor may detect multiple vehicles and their speeds, respectively. Thus, a radar sensor may be used to obtain a speed of each vehicle and its distance to an audio sensor j mounted on a streetlight pole. An audio signal y7(t) may be represented as in Equation (1) as follows:

[0033] Here, n is a number of vehicles, / )(•) is an audio function for a / th audio sensor, s(is a speed of a vehicle z, is a weight of vehicle z, and dij is a distance from vehicle z to an audio sensor j. Vehicle weight has a high impact on noise emitted by the vehicle whenever its speed is high. Then, tire and wind noise dominate. However, when a vehicle’s speed is low, engine and environmental noise dominate. The function / )(•) may be learned with labelled training data at low speed.

[0034] In an embodiment, a radar sensor and an audio sensor at each streetlight pole may record a speed of each passing vehicle and record time-series audio as the vehicle approaches and passes the radar sensor and the audio sensor. When vehicles approach an intersection and a traffic light turns red, the vehicles must reduce their speed. When the detected speed is below lOkm / h, an audio-based machine learning model may classify the vehicles as cars, trucks, buses, etc.

[0035] If there are too many vehicles at an intersection / road section, it may be challenging to recognize each vehicle and separate their individual audio signals for EV detection. An ML learning method may limit the maximum number of vehicles to a threshold and only perform audio-based vehicle recognition and EV detection if the number of vehicles is less than a threshold (e.g., 10). Thus, only a subset of vehicles passing sensors may be sampled.

[0036] A first audio sensor may be located on a first streetlight pole in front of a vehicle. A second audio sensor may be located on a second streetlight pole behind the vehicle. A radar sensor may determine a location of a vehicle. The radar sensor may provide a vehicle’s location to a processor of the first recorded audio and the second recorded audio. ICEVs will show a significant audio difference between the first audio sensor and the second audio sensor due to the location of a combustion engine under a front hood. The audio difference between audio noise at the front and rear of an EV is much smaller than that of an ICEV.

[0037] If a distance of a vehicle to the first audio sensor and the second audio sensor is known, then a vehicle may be determined as an EV or not by comparing the noise which is normalized with respect to the distance. In an embodiment, the first audio sensor may be located on one side of an intersection near a traffic light and the second audio sensor may be located on another side of the traffic light. In an embodiment, a subset of time-series radar and audio data associated with a vehicle changing its speed from lOkmh to a full stop may be used. An audio decibel (dB) level difference between an EV and an ICEV increases as their respective speeds decrease.

[0038] In an embodiment, if an EV emits an additive warning sound (which adds to an audio noise dB level received at an audio sensor), the warning sound may be isolated from the rest of the audio. Additive warning sounds may be filtered from a time series audio input presented to a ML model. The ML model may only be provided with a driving-induced audio noise component and, based thereon, may infer whether a vehicle is an EV or not.

[0039] Whenever there are multiple vehicles significantly contributing to an audio signal received at an audio sensor, successive cancellation methods may be applied to remove audio signals one by one. An audio signal may first be decoded from a closest vehicle, since it may generate the strongest noise, and the audio signal may fade by a power law with respect to a distance to an audio sensor. Then, the audio signals of the closest vehicle are subtracted from the combined signal. Iteration is continued to extract compound audio signal levels of remaining vehicles. Subsequently, an audio noise level of each vehicle i is estimated by Equation (2) as follows: y7(t) is a measured audio signal and y7(t) is an estimated audio signal given the vehicle' s measured speed.

[0040] For each vehicle, a hypothesis test is performed assuming a binary variable (EV or no EV), conditioned on a known vehicle type - car, bus, or truck. For example, a vehicle may be determined as an EV via the minimum distance test of Equation (2) above or by a ML audio test if this is converted to probability.

[0041] In an embodiment, a radar sensor may be used to first recognize a bike based on its speed, which is typically slower than an automobile. Subsequently, audio may be used to classify whether the bike is an EV or not. This allows mapping of electric bike density for a city which will help the city plan charging stations for electric bikes.

[0042] Based on ground truth data (e.g., from a license plate recognition system classifying EVs vs. non EVs in front of a police station), an optimum threshold of data sampling frequency may be determined to meet statistical sufficiency requirements for learning a spatial -temporal density of EVs as well as an upper limit of a number of vehicles / / ., which may be concurrently monitored by Equation (1) above.

[0043] Time series audio data of a vehicle slowing down to below lOkm / h may be required to determine an EV and an ICEV. Vehicles which pass sensors at an intersection at high speed without slowing down may not be classified correctly as an EV or an ICEV. Thus, selective sampling of all slow vehicles in a road section may provide accurate statistics on the EV share on a specific road section of a city. A conventional sampling algorithm (e.g., Gibbs sampling) may be used to infer a probability of spatial -temporal density of EVs in a road section. This may be based on data collected with a time stamp, including a number of EVs detected at each time stamp (e.g., identified from audio data), a number of all vehicles at an instance (e.g., from radar data), and a total number of vehicles across a time period (e.g., from radar data). When the number of vehicles at an instance is too high, EV detection may not be performed due to high noise interference.

[0044] An EV count derived from audio data may be used by a city to forecast EV spatial-temporal charging demand in real time. The data may be used by the city to analyze its CO2 footprint. Similarly, the city may use the data to plan infrastructure to improve road safety based on an inferred density of EVs, density of ICEVs, a number of pedestrians, and a number of bikers.

[0045] FIG. 3 is an exemplary method of determining an EV and an ICEV according to the present disclosure. Step 301 of the method comprises receiving, from a first sensor device at a first location, a first noise level and a first speed of ones of a plurality of vehicles traveling past the first location. Step 303 of the method comprises receiving, from a second sensor device at a second location, a second noise level and a second speed of ones of the plurality of vehicles traveling past the second location. Step 305 of the method comprises classifying, by a processor, one or more of the plurality of vehicles as being either an electric vehicle or an internal-combustion-engine vehicle based on a difference between the first noise level and the second noise level of the vehicle, and further based on the first speed, the second speed, or both the first and second speed of the vehicle.

[0046] In some embodiments, one or more hybrid vehicles may be classified as being operated on a combustion engine or on a battery using techniques described herein.

[0047] Classifying one or more of the plurality of vehicles may include determining if the first speed of the vehicle, the second speed of the vehicle, or both is / are below a predetermined threshold speed. In addition, classifying one or more of the plurality of vehicles may be based on a difference between the first speed and the second speed of the vehicle.

[0048] The sensor devices may comprise a radar sensor to determine the vehicle speeds and a microphone to determine the vehicle noise levels. The first sensor device may be located on a road light fixture. In an embodiment, the first sensor device may be located on a traffic control device, wherein the second sensor device may be located between 25 and 75 meters away from the first sensor device. In an embodiment, a map of traffic flows of electric vehicles and of internal-combustion-engine vehicles may be generated based in part on classifying one or more of the plurality of vehicles. The first noise level, the second noise level, or both of each of the plurality of vehicles may be determined after filtering out vehicle warning noise.

[0049] FIG. 4 is an exemplary system 400 for determining an EV and an ICEV according to the present disclosure. The system 400 comprises a first sensor 401, a second sensor 403, and a processor 405. In an embodiment, the first sensor 401 and the second sensor 403 may communicate with the processor 405 by wired communication, wireless communication, or a combination of both wired communication and wireless communication, where wired communication may be directly wired communication or indirectly wired communication.

[0050] The first sensor 401 may be located at a first location and may be configured to measure a first noise level and a first speed of ones of a plurality of vehicles traveling past the first location. The second sensor 403 may be located at a second location and configured to measure a second noise level and a second speed of ones of the plurality of vehicles traveling past the first location. The processor 405 may be configured to classify one or more of the plurality of vehicles as being either an electric vehicle or an internal-combustionengine vehicle based on a difference between the first noise level and the second noise level of the vehicle, and further based on the first speed, the second speed, or both the first and second speed of the vehicle.

[0051] The first sensor 401 and the second sensor 403 may each comprise a radar sensor to determine a vehicle speed and a microphone to determine the vehicle noise levels. The first sensor 401 may be located on a road light fixture. The first sensor 401 may be located on a traffic control device. The second sensor 403 may be located between 25 and 75 meters away from the first sensor device.

[0052] The processor 405 may be further configured to classify the one or more of the plurality of vehicles in part by determining if the first speed of the vehicle, the second speed of the vehicle, or both is / are below a predetermined threshold speed. The processor 405 may be further configured to classify the one or more of the plurality of vehicles based on a difference between the first speed and the second speed for the vehicle. The processor 405 may be future configured to generate a map of traffic flows of electric vehicles and of internal-combustion-engine vehicles based in part on said classifying.

[0053] If license plate recognition via camera is forbidden, the present disclosure may be used to obtain an EV count as well as an occupant count per car, since radar signals may penetrate a windshield.

[0054] Lithium-ion batteries used in EVs do not perform as well in cold temperatures. Thus, the percentage of Evs on the road may change as temperature changes. The present disclosure may track a density of Evs in different temperatures (e.g., different months) and provide the results to a city for planning purposes.

[0055] Sensors may be mounted on streetlight poles or on a road in marker beacons. The sensors may include an accelerometer to detect vibration of moving parts, an anisotropic magneto-resistive (AMR) detector to sense a vehicle’s electromagnetic field (EMF), a microphone to record an acoustic signature of a vehicle’s operation, a thermal imager to record a heat footprint of a vehicle, and a pressure sensor to monitor air flow.

[0056] In an alternate embodiment, a system includes an outdoor lighting fixture and a sensor to signal a pedestrian of an EV approaching from behind. For example, a lighting installation at a railroad crossing that signals when a train approaches may include a traffic detector to determine traffic type and traffic movement pattern. The system may warn a pedestrian or cyclists of an approaching vehicle from the back. In an alternate embodiment, a system includes an outdoor lighting fixture with an integrated microphone installed in a parking space that monitors an approaching vehicle within a scene and, based on frequency analysis, detects a type of vehicle and adapts the lighting. In parking zones dedicated only to Evs, the light signaling may indicate either an acknowledgement of an authorized vehicle or provide a violation warning if it is an ICEV.

[0057] In an alternate embodiment, a system includes a single-heat-sensing element integrated in a lighting system that monitors a thermal signature of a traversing / standing vehicle to count EV traffic. Furthermore, the traffic may be monitored over time to assist in green accounting to provide an indication of an environmental impact within the scene. Similarly, an estimation of air quality (e.g., pollution-level) of a scene may also be estimated.

[0058] In an alternate embodiment, a system includes an outdoor lighting system to identify unauthorized vehicles to enforcement authorities, reroute traffic (e.g., trafficcalming), or override a current traffic-lighting control scheme for prioritized vehicles (e.g., Evs only). For example, in priority-lanes (e.g., for Evs only), where a flexible fuel vehicle (FFV) is no allowed, the presence of an FFV may be determined.

[0059] In an alternate embodiment, a system detects a stationary ICEV in front of heating, ventilation, and air conditioning (HVAC) vents or closed spaces to notify the ICEV to move.

[0060] In an alternate embodiment, a system monitors a vehicle’s operation to distinguish between an EV and an ICEV due to contrasting vehicle components, overall mechanical composition, and number of components. According to the CALSTART advanced transportation consortium, 70% of an EV’s components may be different from those of an ICEV. An EV has fewer components than an ICEV.

[0061] The NHTSA issued Federal Motor Vehicle Safety Standard (FMVSS) 141, which outlines a minimum sound requirement for Evs and HVs. Thus, auto manufacturers are producing non-ICE vehicles with an additive noise component aimed at signaling the presence of a non-ICE vehicle. In an alternate embodiment, a system detects vehicle type based on the vehicle’s acoustic signature. A high-frequency switching noise made by Evs and HVs at 5 kHz, 10 kHz, and 20kHz, which audibility pedestrians may not hear, may be measured by microphones to classify vehicles. In an embodiment, decision criteria for when to activate EV mapping may include an expected high pollution day and an expected high traffic density day. For example, when a high pollution day is expected, start EV mapping to decide when to regulate a maximum number of combustion-engine vehicles per day that may be allowed into an area (e.g., a city center). For example, when a high traffic density day is expected, start EV mapping to determine whether or not to adjust access policies to an area (e.g., a city, a commuter lane, etc.).

[0062] In another embodiment, two sets of sensors at two different locations may acquire information (e.g., audio) from each vehicle at the same time. Thus, for each vehicle, four measurements may be made (i.e., two measurements from the first set of sensors and two measurements from the second set of sensors, where there are two measurements from a first location and two measurements from a second location). Four measurements may improve EV identification accuracy.

[0063] In an embodiment, audio sensing will utilize a first audio sensor located on a streetlight pole in front of a vehicle and a second audio sensor located on a second streetlight pole behind the vehicle. In an embodiment, a speed sensor (e.g., a radar sensor) may precisely determine a location of the vehicle. The speed sensor provides the vehicle’s location to a first audio artificial intelligence (Al) processor and a second audio Al processor. While an internal-combustion engine may show a significant audio difference between the first audio sensor and a second audio sensor due to the internal-combustion engine being located in a front of the vehicle, the audio difference between audio noise at the front and rear of the vehicle may be much smaller if the vehicle is an EV.

[0064] FIG. 5 shows an illustrative server device 500 that may implement various features and processes as described. The server device 500 may be implemented on any electronic device that runs software applications derived from compiled instructions, including without limitation personal computers, servers, smart phones, media players, electronic tablets, game consoles, email devices, etc. In some implementations, the server device 500 may include one or more processors 502, volatile memory 504, non-volatile memory 506, and one or more peripherals 508. These components may be interconnected by one or more computer buses 510. In an alternate embodiment where real-time service may not be required, computations may be computed using cloud computing. For example, server device 500 may correspond to one or more virtual machine instances within a cloud computer platform.

[0065] Processor(s) 502 may use any known processor technology, including but not limited to graphics processors and multi-core processors. Suitable processors for the execution of a program of instructions may include, by way of example, both general and special purpose microprocessors, and the sole processor or one of multiple processors or cores, of any kind of computer. Bus 510 may be any known internal or external bus technology, including but not limited to ISA, EISA, PCI, PCI Express, NuBus, USB, Serial ATA or FireWire. Volatile memory 504 may include, for example, SDRAM. Processor 502 may receive instructions and data from a read-only memory or a random-access memory or both. The essential elements of a computer may include a processor for executing instructions and one or more memories for storing instructions and data.

[0066] Non-volatile memory 506 may include by way of example semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks and removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. Non-volatile memory 506 may store various computer instructions including operating system instructions 512, communication instructions 514, application instructions 516, and application data 517. Operating system instructions 512 may include instructions for implementing an operating system (e.g., Mac OS®, Windows®, or Linux). The operating system may be multi-user, multiprocessing, multitasking, multithreading, realtime, and the like. Communication instructions 514 may include network communications instructions, for example, software for implementing communication protocols, such as TCP / IP, HTTP, Ethernet, telephony, etc.

[0067] Peripherals 508 may be included within the server device 500 or operatively coupled to communicate with the server device 500. Peripherals 508 may include, for example, network interfaces 518, input devices 520, and storage devices 522. Network interfaces may include, for example an Ethernet or Wi-Fi adapter. Input devices 520 may be any known input device technology, including but not limited to a keyboard (including a virtual keyboard), mouse, trackball, and touch-sensitive pad or display. Storage devices 522 may include one or more mass storage devices for storing data files; such devices include magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and optical disks.

[0068] The system can perform processing, at least in part, via a computer program product, (e.g., in a machine-readable storage device), for execution by, or to control the operation of, data processing apparatus (e.g., a programmable processor, a computer, or multiple computers). Each such program may be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, the programs may be implemented in assembly or machine language. The language may be a compiled or an interpreted language and it may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program may be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network. A computer program may be stored on a storage medium or device (e.g., CD-ROM, hard disk, or magnetic diskette) that is readable by a general or special purpose programmable computer for configuring and operating the computer when the storage medium or device is read by the computer. Processing may also be implemented as a machine-readable storage medium, configured with a computer program, where upon execution, instructions in the computer program cause the computer to operate. The program logic may be run on a physical or virtual processor. The program logic may be run across one or more physical or virtual processors.

[0069] Having described exemplary embodiments of the disclosure, it will now become apparent to one of ordinary skill in the art that other embodiments incorporating their concepts may also be used. The embodiments contained herein should not be limited to disclosed embodiments but rather should be limited only by the spirit and scope of the appended claims. All publications and references cited herein are expressly incorporated herein by reference in their entirety.

[0070] Elements of different embodiments described herein may be combined to form other embodiments not specifically set forth above. Various elements, which are described in the context of a single embodiment, may also be provided separately or in any suitable sub combination. Other embodiments not specifically described herein are also within the scope of the following claims.

[0071] Various embodiments of the concepts, systems, devices, structures and techniques sought to be protected are described herein with reference to the related drawings. Alternative embodiments can be devised without departing from the scope of the concepts, systems, devices, structures and techniques described herein.

[0072] Additionally, the term “exemplary” is used herein to mean “serving as an example, instance, or illustration. Any embodiment or design described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments or designs. The terms “one or more” and “at least one “ are understood to include any integer number greater than or equal to one, i.e., one, two, three, four, etc. The terms "a plurality" are understood to include any integer number greater than or equal to two, i.e., two, three, four, five, etc.

[0073] References in the specification to “one embodiment, “an embodiment,” “an example embodiment,” etc., indicate that the embodiment described can include a particular feature, structure, or characteristic, but every embodiment can include the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

[0074] Use of ordinal terms such as “first,” “second,” “third,” etc., in the claims to modify a claim element does not by itself connote any priority, precedence, or order of one claim element over another or the temporal order in which acts of a method are performed, but are used merely as labels to distinguish one claim element having a certain name from another element having a same name (but for use of the ordinal term) to distinguish the claim elements.

[0075] It is to be understood that the disclosed subject matter is not limited in its application to the details of construction and to the arrangements of the components set forth in the following description or illustrated in the drawings. The disclosed subject matter is capable of other embodiments and of being practiced and carried out in various ways.

[0076] Also, it is to be understood that the phraseology and terminology employed herein are for the purpose of description and should not be regarded as limiting. As such, those skilled in the art will appreciate that the conception, upon which this disclosure is based, may readily be utilized as a basis for the designing of other structures, methods, and systems for carrying out the several purposes of the disclosed subject matter. Therefore, the claims should be regarded as including such equivalent constructions insofar as they do not depart from the spirit and scope of the disclosed subject matter.

[0077] Although the disclosed subject matter has been described and illustrated in the foregoing exemplary embodiments, it is understood that the present disclosure has been made only by way of example, and that numerous changes in the details of implementation of the disclosed subject matter may be made without departing from the spirit and scope of the disclosed subject matter.

Claims

CLAIMS1. A method of detecting vehicles (211, 213), comprising: receiving, from a first sensor device (203) at a first location (205), a first noise level and a first speed of ones of a plurality of vehicles (211, 213) traveling past the first location (205); receiving, from a second sensor device (207) at a second location (209), a second noise level and a second speed of ones of the plurality of vehicles (211, 213) traveling past the second location (209); and classifying, by a processor (405), one or more of the plurality of vehicles (211, 213) as being operated as either an electric vehicle or an internal-combustion-engine vehicle based on a difference between the first noise level and the second noise level of the vehicle, and further based on the first speed, the second speed, or both the first and second speed of the vehicle, wherein an audio noise of intemal-combustion-engine vehicles changes with a speed of the respective one of the internal-combustion-engine vehicles.

2. The method of claim 1, wherein the classifying of the one or more of the plurality of vehicles (211, 213) includes determining if the first speed of the vehicle, the second speed of the vehicle, or both is / are below a predetermined threshold speed.

3. The method of claim 1, wherein the classifying of the one or more of the plurality of vehicles (211, 213) is based on a difference between the first speed and the second speed of the vehicle.

4. The method of claim 1, wherein both of the sensor devices (203, 207) comprise a radar sensor (223) to determine the vehicle speeds and a microphone (217) to determine the vehicle noise levels.

5. The method of claim 1, wherein first sensor device (203) is located on a roadlight or streetlight fixture (205).

6. The method of claim 1, wherein the first sensor device (203) is located on a traffic control device (201), and wherein the second sensor device is located between 25 and 75 meters away from the first sensor device (203).

7. The method of claim 6, further comprising generating a map of traffic flows of electric vehicles and of internal-combustion-engine vehicles based in part on said classifying.

8. The method of claim 1, wherein determining the first noise level, second noise level, or both of each of the plurality of vehicles includes filtering out vehicle warning noise.

9. A system (400) for detecting vehicles, comprising: a first sensor device (401) at a first location configured to measure a first noise level and a first speed of ones of a plurality of vehicles traveling past the first location; a second sensor device (403) at a second location configured to measure a second noise level and a second speed of ones of the plurality of vehicles traveling past the first location; and a processor (405) configured to classify one or more of the plurality of vehicles as being either an electric vehicle or an intemal-combustion-engine vehicle based on a difference between the first noise level and the second noise level of the vehicle, and further based on the first speed, the second speed, or both the first and second speed of the vehicle, wherein an audio noise of internal-combustion-engine vehicles changes with a speed of the respective one of the internal-combustion-engine vehicles.

10. The system of claim 9, wherein the processor (405) is configured to classify the one or more of the plurality of vehicles in part by determining if the first speed of the vehicle, the second speed of the vehicle, or both is / are below a predetermined threshold speed.

11. The system of claim 9, wherein the processor (405) is configured to classify the one or more of the plurality of vehicles based on a difference between the first speed and the second speed for the vehicle.

12. The system of claim 9, wherein both of the sensor devices (401, 403) comprise a radar sensor (223) to determine the vehicle speeds and a microphone (217) to determine the vehicle noise levels.

13. The system of claim 9, wherein first sensor device (401) is located on a roadlight and / or a streetlight fixture (205).

14. The system of claim 9, wherein the first sensor device (401) is located on a traffic control device (201), and wherein the second sensor device (403) is located between 25 and 75 meters away from the first sensor device (401).

15. The system of claim 14, wherein the processor (405) is future configured to generate a map of traffic flows of electric vehicles and of intemal-combustion-engine vehicles based in part on said classifying.

Images