USOO9365213B2
`
`(12) United States Patent
`Stenneth et al.
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 9,365,213 B2
`Jun. 14, 2016
`
`(54) MODE TRANSITION FOR AN AUTONOMOUS
`VEHICLE
`
`(71) Applicant: HERE Global B.V., Veldhoven (NL)
`
`(72) Inventors: Leon Oliver Stenneth, Chicago, IL
`(US); Vladimir Boroditsky,
`Northbrook, IL (US)
`s
`(73) Assignee: HERE Global B.V., Veldhoven (NL)
`-
`(*) Notice:
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 0 days.
`(21) Appl. No.: 14/266,085
`(22) Filed:
`Apr. 30, 2014
`(65)
`Prior Publication Data
`US 2015/0314780 A1
`Nov. 5, 2015
`(51) Int. Cl.
`B60/30/00
`G05D I/00
`B6OW 5O/OO
`(52) U.S. Cl.
`CPC ............. B60W 30/00 (2013.01); G05D 1/0061
`(2013.01); B60W 2050/007 (2013.01); G05D
`2201/0213 (2013.01)
`(58) Field of Classification Search
`CPC ................... B60W 30/182, B60W 2050/0062;
`B60W 2050/0095; B60W 2050/0096; B60W
`30/00, B60W 30/02; B60W 2050/007: G05D
`1/021; G05D 22O1/O213
`See application file for complete search history.
`
`(2006.01)
`(2006.01)
`(2006.01)
`
`(56)
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`8,812, 186 B2 * 8/2014 Oh et al. ......................... TO1/23
`8,880,270 B1 * 1 1/2014 Ferguson et al. ............... TO1/23
`8,954,217 B1 * 2/2015 Montemerlo et al. .......... TO1/26
`2007/02O3617 A1
`8/2007 Haug
`2011/007 1718 A1
`3/2011 Norris et al. .................... TO1/23
`2012,0083959 A1
`4/2012 Dolgov et al.
`2012 0083964 A1
`4/2012 Montemerlo et al.
`2013,0211656 A1
`8, 2013 An et al.
`2013/0253754 A1* 9/2013 Ferguson et al. ............... TO1/28
`2014/0132082 A1* 5, 2014 McGinn et al. ....
`... 307 125
`2014/O136045 A1
`5/2014 Zhu et al. ........................ TO1/23
`2014/0136414 A1* 5/2014 Abhyanker ..................... TO5/44
`2014/0330478 A1* 11/2014 Cullinane et al. ............... TO1/23
`OTHER PUBLICATIONS
`European Search Report cited in EP15160938, mailed Sep. 9, 2015.
`* cited by examiner
`
`Primary Examiner — Rodney Butler
`(74) Attorney, Agent, or Firm — Lempia Summerfield Katz
`LLC
`
`ABSTRACT
`(57)
`An autonomous vehicle may be operable in an autonomous
`mode and a manual mode. A confidence threshold is accessed
`from a database. The confidence threshold may be associated
`with a particular geographic area containing the autonomous
`vehicle. The confidence threshold may be constant for the
`geographic area accessible by the autonomous vehicle. A
`computing device calculates a vehicle confidence level based
`on at least one confidence factor and compares the confidence
`threshold to the vehicle confidence level. The computing
`device generates a driving mode command for a vehicle based
`on the comparison. In one example, the driving mode com
`mand transitions the autonomous vehicle to the autonomous
`mode, if applicable, when the vehicle confidence score
`exceeds the confidence threshold. In one example, the driving
`mode command transitions the autonomous vehicle to the
`manual mode, if applicable, when the vehicle confidence
`score does not exceed the confidence threshold.
`
`8,078,349 B1
`8,718,861 B1*
`
`12/2011 Prada Gomez et al.
`5/2014 Montemerlo et al. .......... TO1/26
`
`19 Claims, 10 Drawing Sheets
`
`Sveticle confidence level exceed
`cificiece tres
`
`: ...
`
`
`
`
`
`S23
`
`Geneate autonomous driving
`{Orman
`
`IPR2025-00943
`Tesla EX1024 Page 1
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 1 of 10
`
`US 9,365,213 B2
`
`2
`
`WA EWE.OER
`
`
`
`
`
`NEWORK
`27
`
`
`
`
`
`
`
`
`
`computing
`DEVICE 122
`
`G.
`
`IPR2025-00943
`Tesla EX1024 Page 2
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 2 of 10
`
`US 9,365,213 B2
`
`
`
`G. 2
`
`IPR2025-00943
`Tesla EX1024 Page 3
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 3 of 10
`
`US 9,365,213 B2
`
`
`
`32
`
`
`
`G. 3A
`
`IPR2025-00943
`Tesla EX1024 Page 4
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 4 of 10
`
`US 9,365,213 B2
`
`G 4A
`
`138
`
`2OO
`
`
`
`
`
`39 -A- 5 2
`
`IPR2025-00943
`Tesla EX1024 Page 5
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 5 of 10
`
`US 9,365,213 B2
`
`FG. A
`
`4.
`
`
`
`IPR2025-00943
`Tesla EX1024 Page 6
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 6 of 10
`
`US 9,365,213 B2
`
`
`
`
`
`343,
`
`IPR2025-00943
`Tesla EX1024 Page 7
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 7 of 10
`
`US 9,365,213 B2
`
`s
`
`y
`
`
`
`SERVER
`
`{Ovy MCA ON
`NERACE
`
`G. 7
`
`IPR2025-00943
`Tesla EX1024 Page 8
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 8 of 10
`
`US 9,365,213 B2
`
`S. f
`
`S103
`
`Access a confidence threshoid.
`
`Cacuate a vehicle confidence score based of sensor data.
`
`SOS
`
`Performing a comparison of the confidence threshoid to the vehicle
`confidence eve,
`
`S. Of
`
`l -- Y.
`Generate an autonomous driving command for a vehicle based of the
`comparison.
`
`G. 8
`
`IPR2025-00943
`Tesla EX1024 Page 9
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 9 of 10
`
`US 9,365,213 B2
`
`
`
`22
`
`WO3
`
`EVCE
`
`COMMUNICATION
`NTERFACE
`
`OSO
`CRCRY
`
`SPAY
`
`FG, 9
`
`IPR2025-00943
`Tesla EX1024 Page 10
`
`

`

`U.S. Patent
`
`Jun. 14, 2016
`
`Sheet 10 of 10
`
`US 9,365,213 B2
`
`$205
`
`Does vehicle confidence level exceed
`Cofidence thresioid
`
`S207
`
`
`
`Yes
`
`f
`
`
`
`S29
`l
`
`Generate autonomous driving
`coard.
`
`Generate ran a driving
`command.
`
`sessssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssss
`
`G. O.
`
`IPR2025-00943
`Tesla EX1024 Page 11
`
`

`

`1.
`MODE TRANSTION FOR AN AUTONOMOUS
`VEHICLE
`
`US 9,365,213 B2
`
`FIELD
`
`The following disclosure relates to the transition of modes
`of an autonomous vehicle, or more particularly, confidence
`levels for controlling the transition of modes of an autono
`mous vehicle.
`
`BACKGROUND
`
`The term autonomous vehicle refers to a vehicle including
`automated mechanisms for performing one or more human
`operated aspects of vehicle control. As autonomous vehicles
`are adopted, several benefits may be realized. Vehicle colli
`sions may be reduced because computers can perform driving
`tasks more consistently and make fewer errors than human
`operators. Traffic congestion may be alleviated because
`autonomous vehicles observe specified gaps between
`vehicles, preventing stop and go traffic. The reduced traffic
`and increased safety may lead to higher speed limits.
`Autonomous vehicles may allow drivers to focus their
`attention elsewhere, such as working on a laptop, talking on a
`phone, or sleeping. Impaired people that may otherwise be
`unable to drive may be able to operate an autonomous vehicle.
`Parking options in urban errors may be improved because
`autonomous vehicles may drop off passengers and then park
`in a more remote location.
`However, autonomous vehicles may be operable only on
`certain roads or certain types of roads. Autonomous vehicle
`passengers may need to operate the vehicles in some areas.
`Challenges remain in providing transitions between auto
`mated mode and human operated mode.
`
`SUMMARY
`
`10
`
`15
`
`25
`
`30
`
`35
`
`In one embodiment, an autonomous vehicle may be oper
`able in an autonomous mode and a manual mode. A confi
`dence threshold is accessed from a database. The confidence
`threshold may be associated with a particular geographic area
`containing the autonomous vehicle. The confidence threshold
`may be constant for the geographic area accessible by the
`autonomous vehicle. A computing device calculates a vehicle
`confidence level based on at least one confidence factor and
`compares the confidence threshold to the vehicle confidence
`level. The computing device generates a driving mode com
`mand for a vehicle based on the comparison. In one example,
`the driving mode command transitions the autonomous
`vehicle to the autonomous mode, if applicable, when the
`vehicle confidencescore exceeds the confidence threshold. In
`one example, the driving mode command transitions the
`autonomous vehicle to the manual mode, if applicable, when
`the vehicle confidence score does not exceed the confidence
`threshold.
`
`40
`
`45
`
`50
`
`55
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`Exemplary embodiments of the present invention are
`described herein with reference to the following drawings.
`FIG. 1 illustrates an example system for mode transition
`for an autonomous vehicle.
`FIG. 2 illustrates example sensors for vehicle confidence.
`FIG. 3A illustrates example set of geographic Zones.
`FIG. 3B illustrates another example set of geographic
`ZOS.
`FIG. 4A illustrates an example confidence score display.
`
`60
`
`65
`
`2
`FIG. 4B illustrates another example confidence score dis
`play.
`FIG.5A illustrates an example look up table for confidence
`thresholds.
`FIG. 5B illustrates another example look up table for
`vehicle confidence scores.
`FIG. 6A illustrates an example grid neighborhood.
`FIG. 6B illustrates another example grid neighborhood.
`FIG. 7 illustrates an exemplary server of the system of FIG.
`1.
`FIG. 8 illustrates example flowchart for mode transition for
`an autonomous vehicle.
`FIG. 9 illustrates an exemplary computing device of the
`system of FIG. 1.
`FIG. 10 illustrates another example for mode transition for
`an autonomous vehicle.
`
`DETAILED DESCRIPTION
`
`Some autonomous vehicles may include a completely driv
`erless mode in which no passengers are onboard. These
`autonomous vehicles may park themselves or move cargo
`between locations without a human operator. Autonomous
`vehicles may include multiple modes and transition between
`the modes. A highly assisted driving mode may not com
`pletely replace the human operator. In the highly assisted
`driving mode, the vehicle may perform some driving func
`tions and the human operator may perform some driving
`functions. The vehicles may also be driven in a manual mode
`in which the human operator exercises a degree of control
`over the movement of the vehicle. The vehicles may also
`include the completely driverless mode. Other levels of auto
`mation are possible.
`The transition between driving modes may be automatic or
`manual. The mode transition may be between the driverless
`mode and the manual mode, between the highly assisted
`driving mode and the manual mode, or between the highly
`assisted driving mode and the automatic mode. A manual
`transition may be controlled by the press of a button, lever, or
`touchscreen within the vehicle. The automatic transition may
`be made by the vehicle. The transition may be based on
`confidence levels.
`The vehicle may calculate a vehicle confidence level based
`on the current external conditions the vehicle is facing. The
`vehicle confidence level may be compared to a threshold
`value to determine the appropriate mode for the vehicle.
`Alternatively, the vehicle confidence level may be compared
`to a geographic confidence level that describes the requisite
`confidence for the current geographic area.
`FIG. 1 illustrates an example system 120 for mode transi
`tion for an autonomous vehicle. The system 120 includes a
`developer system 121, one or more computing devices 122, a
`workstation 128, and a network 127. Additional, different, or
`fewer components may be provided. For example, many com
`puting devices 122 and/or workstations 128 connect with the
`network 127. The developer system 121 includes a server 125
`and a database 123. The developer system 121 may include
`computer systems and networks of a system operator.
`The computing device 122 may be carried by a vehicle 124.
`The computing device 122 may be a specialized autonomous
`driving computer. The computing device 122 may calculate a
`vehicle confidence level based on at least one confidence
`factor. The confidence factors may be based on sensor data
`collected at the vehicle, environmental data received through
`the network 127, or responsiveness of the vehicle 124. Alter
`natively, the computing device 122 may report sensor data to
`the server 125, which calculates the vehicle confidence level.
`
`IPR2025-00943
`Tesla EX1024 Page 12
`
`

`

`US 9,365,213 B2
`
`10
`
`15
`
`3
`The computing device 122 or the server 125 may perform
`a comparison to the vehicle confidence value to a confidence
`threshold. The computing device 122 generates a driving
`mode command for the vehicle 124 based on the comparison.
`When the vehicle confidence exceeds the confidence thresh
`old, the computing device 122 determines that the conditions
`are suitable for autonomous driving. Thus, the driving mode
`command includes data for a transition to or to remain in the
`autonomous mode for the vehicle 124. When the vehicle
`confidence is less than the confidence threshold, the comput
`ing device 122 determines that the conditions are unsuitable
`for autonomous driving. Thus, the driving mode command
`includes data for a transition to or to remain in the manual
`mode for vehicle 124.
`The confidence threshold may be stored in database 123.
`The confidence threshold may be associated with a geo
`graphic area. The confidence threshold may be constant
`through the geographic area. Thus, some vehicles may use the
`same confidence threshold at all times. In other example, the
`confidence threshold may change as the vehicles travel.
`The computing device 122 may be a device that is in
`communication with the autonomous driving computer built
`in to the vehicle 124. The computing device 122 may perform
`the confidence comparison and send the results to the autono
`mous driving computer. The computing device 122 may be a
`mobile device Such as a Smart phone, a mobile phone, a
`personal digital assistant (PDA), a tablet computer, a note
`book computer, a navigation device, and/or any other known
`or later developed portable or mobile computing device.
`The optional workstation 128 is a general purpose com
`30
`puter including programming specialized for the following
`embodiments. For example, the workstation 128 may receive
`user inputs for the confidence threshold or the confidence
`factors.
`The developer system 121, the workstation 128, and the
`computing device 122 are coupled with the network 127. The
`phrase “coupled with is defined to mean directly connected
`to or indirectly connected through one or more intermediate
`components. Such intermediate components may include
`hardware and/or software-based components.
`The computing resources may be divided between the
`server 125 and the computing device 122. In some embodi
`ments, the server 125 performs a majority of the processing
`for calculating the vehicle confidence value and the compari
`son with the confidence threshold. In other embodiments, the
`computing device 122 or the workstation 128 performs a
`majority of the processing. In addition, the processing is
`divided substantially evenly between the server 125 and the
`computing device 122 or workstation 128.
`FIG. 2 illustrates example sensors for vehicle confidence.
`The sensors may be organized in to overlapping or non
`overlapping categories. One classification of sensors includes
`internal vehicle sensors (e.g., sensors 111 and 113), external
`vehicle sensors (e.g., sensors 115 and 119), and stationary
`sensors (e.g., sensors 117). Another classification of sensors
`may include driving sensors (e.g., sensors 111,113 and 115),
`road status sensors (e.g., sensors 115, 117, and 119), and
`parking sensors (e.g., sensors 111 and 113).
`Engine sensors 111 may include throttle sensor that mea
`Sures a position of a throttle of the engine or a position of an
`60
`accelerator pedal, a brake Senor that measures a position of a
`braking mechanism or a brake pedal, or a speed sensor that
`measures a speed of the engine or a speed of the vehicle
`wheels. Vehicle sensors 113 may include a steering wheel
`angle sensor, a speedometer sensor, or a tachometer sensor.
`When the vehicle 125 is traveling higher than a predeter
`mined speed the vehicle confidence level may be raised. In
`
`50
`
`4
`another example, when the brake is pressed at a predeter
`mined pressure, the vehicle confidence level may be raised. In
`another example, when the steering wheel is turned to a
`predetermined angle, the vehicle confidence level may be
`raised.
`The external vehicle sensor 115 may be a camera, a light
`detection and ranging (LIDAR) sensor, a radar sensor, or an
`ultrasonic sensor. The external vehicle sensor 115 may deter
`mine road status Such as the shape or turns of the road, the
`existence of speed bumps, the existence of pot holes, the
`wetness of the road, or the existence or ice, Snow, or slush.
`The secondary vehicle sensor 119 may be a camera, a
`LIDAR sensor, a radar sensor, or an ultrasonic sensor. The
`secondary vehicle sensor 119 may mounted on another
`vehicle. The secondary vehicle sensor 119 may be a backup
`sensor for detecting the speed of the vehicle 124. The second
`ary vehicle sensor 119 may detect the weather in the vicinity
`of the vehicle 124.
`The stationary sensor 117 may be a camera, a LIDAR
`sensor, a radar sensor, or an ultrasonic sensor. The stationary
`sensor 117 may be a backup sensor for detecting the speed of
`the vehicle 124. The stationary sensor 117 may detect the
`weather in the vicinity of the vehicle 124. The stationary
`sensor 117 may detect traffic levels of the roadway.
`One or more of the sensors may be weather sensors such as
`wiperblade sensors, rain sensors, temperature sensors, baro
`metric sensors, or other types of sensors related to the
`weather. Rain, sleet, Snow, fog, or barometric changes may be
`indicative of more hazardous driving conditions and may
`lower the vehicle confidence level. Temperatures below
`freezing, especially in combination with other weather con
`ditions, may be indicative of more hazardous driving condi
`tions and may lower the vehicle confidence level.
`One or more of the sensors may include infotainment sen
`sors. The infotainment sensor may be an AM/FM sensor or a
`video sensor. The infotainment sensor may receive traffic
`message channel broadcasts. Traffic data in the broadcasts
`may be used to modify the vehicle confidence level. Also,
`when the radio is played above a certain level or video is being
`watched, the vehicle confidence level may be raised so that
`the vehicle enters autonomous mode, which is preferred to a
`distracted manual driver.
`One or more of the sensors may include parking sensors.
`The parking sensor category may include any combination of
`seatbelt sensors, door sensors, or a gear shift sensor. Various
`indicators from the parking sensors may be used to determine
`that the vehicle 124 is parking or moving away from a parking
`spot. The indicators may be fastening seatbelts, closings
`doors, or shifting to a higher gear. In one example, the vehicle
`confidence level is raised as the vehicle 124 is successfully
`moved from a parking spot.
`One or more of the vehicle sensors 113 may detect a driver.
`The sensor 113 may be an eye gaze sensor or a video sensor
`that determines whether the driver is asleep or has been nod
`ding off or whether the driver is paying attention to the road
`ahead. The sensor 113 may be a weight sensor that determines
`whetheran adult sized driver is in the driver's seat. The sensor
`113 may be a breathalyzer that detects the breath of the driver
`to estimate the blood alcohol content of the driver. When the
`driver is drowsy, nonexistent, or impaired as indicated by the
`sensor 113, the vehicle confidence value may be raised to
`increase the likelihood that the vehicle 124 enters the autono
`mous mode.
`The sensors data may be sent to an automotive cloud stored
`at database 123 or at the server 125 through network 127. The
`vehicles 124 may submit and download data from the auto
`motive cloud. The data from other vehicles stored in the
`
`25
`
`35
`
`40
`
`45
`
`55
`
`65
`
`IPR2025-00943
`Tesla EX1024 Page 13
`
`

`

`US 9,365,213 B2
`
`5
`automotive cloud may be used to modify the vehicle confi
`dence level or define the confidence threshold.
`FIG. 3A illustrates example set of geographic Zones 133.
`The world 132 or any other geographic area may be divided
`into geographic Zones 133. In one example, the geographic
`Zones 133 are defined according to the amateur (ham) radio
`grids that divide the earth into 324 Zones (18 divisions in the
`longitude direction and 18 divisions in the latitude direction).
`The geographic Zones 133 may be designated by two charac
`ter alphanumeric codes.
`The geographic Zones 133 may be divided into other sized
`intervals. In one example, the geographic Zones 133 may
`encompass a predetermined geographic space (e.g., 1 Square
`mile, 1 Square kilometer, 10 Square miles or another value).
`The geographic Zones 133 may be defined according to juris
`dictions or governmental authority (e.g., by county, by city,
`by state, or by country). The government authority may set the
`confidence threshold or a geographic confidence level.
`FIG. 3B illustrates another example set of geographic
`Zones. The geographic Zones may represent individual path
`segments or sets of path segments. For example, a highway
`from point A to B may be in a single geographic Zone. FIG.3B
`illustrates geographic zones 133a and 133b, which are of
`unequal size.
`The sensors may include global positioning system (GPS)
`or the computing device 122 includes one or more detectors
`or sensors as a positioning system built or embedded into or
`within the interior of the computing device 122. The comput
`ing device 122 receives location data for geographic position
`from the positioning system.
`The computing device 122 receives location data for the
`vehicle 124 in the geographic area. The computing device 122
`compares the location data to the geographic Zones 133 to
`identify the current geographic Zone of the vehicle 124. For
`example, the computing device 122 may access the database
`123 is according to the location data to receive the confidence
`threshold for the corresponding geographic Zone or grid.
`Alternatively, the location data may be sent to the server 125,
`which determines the appropriate geographic Zone.
`The confidence thresholds for the geographic Zones may be
`maintained by the server 125 or another device. The confi
`dence thresholds may be set at a default value and increased
`or decreased depending on the conditions found within the
`geographic Zones. Factors that impact the confidence thresh
`olds include weather, traffic, the type of road and road fea
`tures, data from road sensors, vehicle data from recent
`vehicles in the Zone, blind spots in the Zone, incident statistics
`for the Zone, and a mobile object flow rate for the Zone.
`The weather in a geographic Zone may be determined
`based on weather sensors (e.g., temperature sensors, rain
`sensors, or barometric sensors) physically located with the
`geographic Zone. The weather may be determined from a
`weather service. Weather conditions such as rain, Snow, and
`fog may increase the confidence threshold Such that vehicles
`must have higher confidence levels in order to enter autono
`mous mode in the geographic Zone.
`The traffic in the geographic Zone may be determined by a
`traffic service Such as traffic message channel. The traffic may
`be determined based on the speeds of vehicles in the geo
`graphic Zones. In one example, when traffic is increased in a
`geographic Zone, the threshold is raised because slowertraffic
`is easier for the autonomous mode to control the vehicle or
`because the autonomous mode is better adept at spacing
`vehicle to eliminate the traffic. In another example, when
`traffic is increased in a geographic Zone, the confidence
`threshold is lowered so that vehicles are more likely to be in
`manual mode. Road sensors may also report traffic levels.
`
`40
`
`45
`
`6
`The type of road and road features may include functional
`class, which is discussed in more detail below, speed limits,
`road quality, locations of stop signs and lights, or other road
`features. As road features are more numerous, roads may be
`more difficult for the autonomous to manage, and the confi
`dence threshold is increased.
`Vehicle data from recent vehicles in the Zone in the last
`predetermined time may impact the Zones geographic levels.
`The vehicle data may include any of the vehicle sensors
`discussed above. Reported blind spots in the Zone may
`describe places where autonomous mode is more reliable
`than manual model because automated systems are not
`affected by driver blind spots. The confidence threshold may
`be modified based on a mobile object flow rate for the Zone.
`The mobile object flow rate may describe a number of mobile
`objects (e.g., deer, kids, pets, or other objects) that is expected
`to flow through the grid. The number of mobile objects may
`depend on the relative distances to sources of mobile objects
`(e.g., parks, Zoos, wooded areas, playgrounds, or other areas).
`The confidence threshold may also be affected by accident
`statistics or other incidents in the geographic Zone. The inci
`dent reports may be in near real time, which may be included
`in traffic data. The incident reports may be a statistic kept over
`alonger period of time such as the last month, year, or all time
`available. Higher incident Zones could be assigned a higher
`confidence threshold so that more vehicles are in autonomous
`mode and may avoid incident. In another example, higher
`incident Zones could be assigned a lower confidence thresh
`old if it is determined a substantial portion of the incident
`involve vehicles in autonomous mode.
`The confidence threshold may be a function of time. Each
`geographic Zone may be assigned multiple confidence thresh
`old based on any of the above factor for multiple time inter
`vals or time epochs. Each geographic Zone may be associated
`with a time matrix stored in the database 123. The matrix may
`include data for each specific time epoch, or the matrix may
`include multiple layers or divisions for multiple time epochs.
`The time epoch corresponds to a time of day, day of week,
`and/or day of the year. The size of the repeating time epochs
`may be configurable by the developer. Example sizes include
`15 minutes, 30 minutes, 1 hour, or another value. Example
`time epochs include 5:00 pm-5:15 pm on a weekday, 10:00
`am-11:00am on a Saturday, and 2:00pm-4:00 pm on the day
`before Thanksgiving. In the example of 15 minute epochs, the
`confidence threshold may beformatted into a 96-dimensional
`vector for each cell of the matrix, in which each of the 96
`components describe speed data for a different 15 minute
`epoch. The size of matrix is a function of the size, or quantity,
`of the time epochs.
`FIG. 4A illustrates an example confidence score display
`137. The display 137 may be part of computing device 122
`(e.g., in dash display, or a mobile device). The display 137
`may illustrate a numeric value that represents the current
`vehicle score of the vehicle 124. The vehicle score may fluc
`tuate according to sensor data received at the computing
`device 122. In one example, the display 137 includes an
`absolute value for the vehicle confidence level. In another
`example, the display 137 includes a differential value that
`represented the difference between the vehicle confidence
`level and the threshold. When the vehicle confidence level
`exceeds the threshold, a positive value is displayed, and when
`the vehicle confidence level is lower that the threshold, a
`negative value is displayed.
`FIG. 4B illustrates an example gradient confidence score
`display 138. The display 138 may include a positive region to
`indicate the vehicle confidence level exceeds the threshold.
`The display 138 may include a negative region to indicate the
`
`10
`
`15
`
`25
`
`30
`
`35
`
`50
`
`55
`
`60
`
`65
`
`IPR2025-00943
`Tesla EX1024 Page 14
`
`

`

`US 9,365,213 B2
`
`10
`
`15
`
`7
`vehicle confidence level exceeds is less than the threshold.
`The display 138 may include a neutral within a predetermined
`range (e.g., 5%) of the threshold value. Indicia (e.g., arrow
`139) may indicate the current vehicle confidence value.
`The positive portion of the display 138 may be broken 5
`down into multiple ranges. Each range may correspond to a
`different set of autonomous vehicle functions. A first range,
`for example from the neutral threshold value to a first tier
`threshold (e.g., 10 absolute points higher or 10% higher than
`the threshold value), the autonomous driving computer may
`include low risk driving functions. Low risk driving functions
`may include windshield wiper control, headlight operation,
`and cruise control.
`A second range, for example from the first tier threshold
`value to a second tier threshold (e.g., 20 absolute points
`higher or 20% higher than the threshold value), the autono
`mous driving computer may include medium risk driving
`functions. Medium risk driving functions may include speed
`control, gear shifting or lane departure prevention. Speed 20
`control may prevent the vehicle from moving within a prede
`termined range of the vehicle ahead or from exceeding the
`speed limit. Gear shifting may control the transmission of the
`vehicle to Supplement speed control. Lane departure preven
`tion may make minor steering adjustments based on the lane 2s
`of travel through detect roadway lines or location tracking.
`A third range, for example from the second threshold value
`to a third tier threshold (e.g., 50 absolute points higher, 50%
`higher than the threshold value, or an open ended maximum
`value), the autonomous driving computer may include high 30
`risk driving functions. High risk driving functions may be full
`automated driving including steering, speed control, gear
`shifting, and braking.
`FIG. 5A illustrates an example look up table 145 for con
`fidence thresholds or the confidence scores of the geographic 35
`Zones. The lookup table may be stored at the server 125, the
`database 123, or at the computing device 122. The table 145
`illustrates a correlation between functional class of paths and
`the corresponding threshold value for the geographic Zone.
`In one example, the geographic Zone relates to individual 40
`road segments, and the threshold value for the road segments
`is based on functional class and any of the other factors above.
`In another example, the geographic Zone includes paths of
`multiple functional classifications. The Zone's geographic
`threshold may be multiplied by a multiplier depending on the 4s
`functional classes of the individual path segments. The mul
`tipliers may be selected Such that Smaller roads (higher func
`tional class) have a higher multiplier, which results in a higher
`threshold value and reduced the amount of autonomous driv
`ing available on that path segment.
`Table 1 lists example classification systems that may be
`assigned numeric values for functional class.
`
`8
`roads, collector roads, and local roads. The functional classi
`fications of roads balance between accessibility and speed.
`An arterial road has low accessibility but is the fastest mode of
`travel between two points. Arterial roads are typically used
`for long distance travel. Collector roads connectarterial roads
`to local roads. Collector roads are more accessible and slower
`than arterial roads. Local roads are accessible to individual
`homes and business. Local roads are the most accessible and
`slowest type of road.
`An example of a complex functional classification system
`is the urban classification system. Interstates include high
`speed and controlled access roads that span long distances.
`The arterial roads are divided into principle arteries and minor
`arteries according to size. The collector roads are divided into
`major collectors and minor collectors according to size.
`Another example functional classification system divides
`long distance roads by type of road or the entity in control of
`the highway. The functional classification system includes
`interstate expressways, federal highways, state highways,
`local highways, and local access roads. Another functional
`classification system uses the highway tag system in the Open
`Street Map (OSM) system. The functional classification
`includes motorways, trunk roads, primary roads, secondary
`roads, tertiary roads, and residential roads.
`Alternatively, the database 123 may list road width or lane
`quantities. The server 125 may access the database for the
`road width according to the geographic location reported by
`the computing device 122 and assign a multiplier to the
`threshold value of the Zone based on road width or the number
`of lanes.
`FIG. 5B illustrates another example look up table 147 for
`vehicle confidence score. As discussed above, the confidence
`score for a vehicle may dependent on data received from a
`variety of sensors, which may be classified by sensor cat
`egory. In addition, to the sensor data returned by the sensors,
`the vehicle confidence scores may also depend on the quality
`of the data from the sensors.