IP Library Granted Patent US 10,482,003
Granted Patent B1
US 10,482,003 · App. 16/185,088 · Granted Nov 19, 2019

Method and system for modifying a control unit of an autonomous car

Inventors: András Bondor (Martonvásár, HU); László Kishonti (Budapest, HU)
Assignee: AImotive Kft.
G06F11/3684G05D1/0088G06F17/5095G05D2201/0213G06K9/00791
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,482,003
App. No.
16/185,088
Granted
Nov 19, 2019
Kind
B1
Abstract

Sensor data is recorded during testing of a control unit of an autonomous car in an autonomous driving mode. In response to being triggered by a disengagement event, recorded sensor data from an analysis time window prior to the disengagement event is logged. The logged sensor data is processed in a simulator using the control unit to control a virtual autonomous car to identify the inadequately operating component of the control unit. The component may be modified and retested until the disengagement event is avoided in the simulator. The control unit with the modified component may then be used to record new sensor data during testing of the control unit in the autonomous driving mode, and the simulation repeated on the new data.

Claims (24)

1. A method for modifying a control unit of an autonomous car, the autonomous car being equipped with basic sensors adapted for recording sensor data based on which a fusion model of objects and an environment surrounding the autonomous car can be generated, wherein the sensor data recorded by the basic sensors enables the control unit to control the autonomous car in autonomous driving mode, the method comprising:

while the control unit is in a first state, recording the sensor data by the basic sensors during an autonomous test action performed in autonomous driving mode of the autonomous car,

in response to being triggered by a disengagement event, logging disengagement related data comprising the sensor data recorded by the basic sensors in an analysis time window before the disengagement event,

generating a virtual scenario in a simulator apparatus for the analysis time window based on the disengagement related data, wherein the virtual scenario comprises a virtual counterpart of the autonomous car having virtual counterparts of the basic sensors, and performing a first virtual autonomous run according to the autonomous test action with the virtual counterpart of the autonomous car using the first state of the control unit in the simulator apparatus, so that the first virtual autonomous run leads to the disengagement event in the virtual scenario,

identifying based on the virtual scenario an inadequately operating component of the control unit, wherein the disengagement event has been caused by inadequate operation of the inadequately operating component of the control unit, and

generating a second state of the control unit based on the first state of the control unit by modifying the inadequately operating component of the control unit, performing a second virtual autonomous run according to the autonomous test action with the virtual counterpart of the autonomous car using the second state of the control unit in the simulator apparatus, checking whether the disengagement event is avoided in the second virtual autonomous run, and storing the modified component of the control unit if the disengagement event is avoided in the second virtual autonomous run.

2. The method according to claim 1 , further comprising applying validation data in generating the virtual scenario for validating a validation scenario part of the virtual scenario based on comparing a validation model of the validation scenario part to the fusion model of the validation scenario part for at least a part of the analysis time window, and the inadequately operating component of the control unit is identified by means of the comparison of the validation model of the validation scenario part and the fusion model of the validation scenario part for at least the part of the analysis time window.

3. The method according to claim 2 , characterized in that the autonomous car is equipped with a validator sensor adapted for recording the validation data.

4. The method according to claim 3 , characterized in that the validator sensor is different from the basic sensors.

5. The method according to claim 4 , characterized in that the validator sensor is a lidar device.

6. The method according to claim 5 , characterized in that a distance estimation component of the control unit is identified as the inadequately operating component of the control unit, and the distance estimation component of the control unit is modified for generating the second state of the control unit based on the first state of the control unit.

7. The method according to claim 1 , characterized in that, if the fusion model generated by means of the basic sensors in the analysis time window is found to be acceptable in the analysis time window, after performing the second virtual autonomous run, a third virtual autonomous run is performed according to the autonomous test action with the virtual counterpart of the autonomous car giving the disengagement related data to virtual counterparts of the basic sensors of the virtual counterpart of the autonomous car using the second state of the control unit in the simulator apparatus.

8. A system for modifying a control unit of an autonomous car, the system comprising a computer connected to the control unit and including instructions that, when executed by the computer, cause the computer to:

while the control unit is in a first state, record sensor data from basic sensors of the autonomous car during an autonomous test action performed in autonomous driving mode of the autonomous car,

in response to being triggered by a disengagement event, log disengagement related data comprising the sensor data recorded in an analysis time window before the disengagement event,

generate a virtual scenario in a simulator apparatus for the analysis time window based on the disengagement related data, wherein the virtual scenario comprises a virtual counterpart of the autonomous car having virtual counterparts of the basic sensors, and perform a first virtual autonomous run according to the autonomous test action with the virtual counterpart of the autonomous car using the first state of the control unit in the simulator apparatus, so that a first virtual autonomous run leads to the disengagement event in the virtual scenario,

identify, based on the virtual scenario, an inadequately operating component of the control unit, wherein the disengagement event has been caused by inadequate operation of the inadequately operating component of the control unit, and

generate a second state of the control unit based on the first state of the control unit by modifying the inadequately operating component of the control unit, perform a second virtual autonomous run according to the autonomous test action with the virtual counterpart of the autonomous car using the second state of the control unit in the simulator apparatus, check whether the disengagement event is avoided in the second virtual autonomous run, and store the modified component of the control unit if the disengagement event is avoided in the second virtual autonomous run.

9. A computer program product comprising a non-transitory computer readable medium storing instructions which, when executed by a computer, cause the computer to:

while the control unit is in a first state, record sensor data from basic sensors of the autonomous car during an autonomous test action performed in autonomous driving mode of the autonomous car,

in response to being triggered by a disengagement event, log disengagement related data comprising the sensor data recorded in an analysis time window before the disengagement event,

generate a virtual scenario in a simulator apparatus for the analysis time window based on the disengagement related data, wherein the virtual scenario comprises a virtual counterpart of the autonomous car having virtual counterparts of the basic sensors, and perform a first virtual autonomous run according to the autonomous test action with the virtual counterpart of the autonomous car using the first state of the control unit in the simulator apparatus, so that a first virtual autonomous run leads to the disengagement event in the virtual scenario,

identify, based on the virtual scenario, an inadequately operating component of the control unit, wherein the disengagement event has been caused by inadequate operation of the inadequately operating component of the control unit, and

generate a second state of the control unit based on the first state of the control unit by modifying the inadequately operating component of the control unit, perform a second virtual autonomous run according to the autonomous test action with the virtual counterpart of the autonomous car using the second state of the control unit in the simulator apparatus, check whether the disengagement event is avoided in the second virtual autonomous run, and store the modified component of the control unit if the disengagement event is avoided in the second virtual autonomous run.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 8, 2019
From: BONDER, ANDRÁS; KISHONTI, LÁSZLÓ
To: AIMOTIVE KFT.
Reel/Frame 048280/0065 →
Cited By (8)
US 12,254,763 US 12,304,510 US 12,358,518 US 12,374,171 US 12,423,759 US 12,497,028 US 12,555,417 US 12,716,733