Introspective competence modeling for AV decision making

Patent metadata

Field	Value
Publication number	US-11307585-B2
Application number	US-201916668584-A
Country	US
Kind code	B2
Filing date	Oct 30, 2019
Priority date	Oct 30, 2019
Publication date	Apr 19, 2022
Grant date	Apr 19, 2022

Abstract

A first method includes detecting, based on sensor data, an environment state; selecting an action based on the environment state; determining an autonomy level associated with the environment state and the action; and performing the action according to the autonomy level. The autonomy level can be selected based at least on an autonomy model and a feedback model. A second method includes calculating, by solving an extended Stochastic Shortest Path (SSP) problem, a policy for solving a task. The policy can map environment states and autonomy levels to actions and autonomy levels. Calculating the policy can include generating plans that operate across multiple levels of autonomy.

First claim

What is claimed is:

1. A method of autonomous driving by an autonomous vehicle (AV), comprising:
detecting, based on sensor data, an environment state;
selecting an action based on the environment state;
determining an autonomy level associated with the environment state and the action,
wherein the autonomy level indicates an extent to which the action is to be performed autonomously by the AV,
wherein the AV determines respective autonomy levels for actions to be performed,
wherein the autonomy level is selected based at least on an autonomy model and a feedback model, and
wherein the feedback model models feedback types received from a human with respect to the action, how costly each type of feedback is, and how likely is each type of feedback to be received from the human; and

performing the action according to the autonomy level, wherein performing the action according to the autonomy level comprises:
in response to determining that the autonomy level indicates “verified autonomy” and to determining that the AV is being monitored:
performing the action; and
in response to receiving an override signal while the action is being performed, stopping the action and switching to a manual operation mode of the AV.

2. The method of claim 1 , wherein the autonomy level is selected from a set comprising a first autonomy level indicating “no autonomy”, a second autonomy level indicating “verified autonomy”, a third autonomy level indicating “supervised autonomy”, and a fourth autonomy level indicating “unsupervised autonomy”.

3. The method of claim 2 ,
wherein the autonomy level is the second autonomy level indicating “verified autonomy”, and wherein performing the action according to the autonomy level comprises:
receiving, for the action, an approval feedback signal or a disapproval feedback signal.

4. The method of claim 3 , wherein performing the action according to the autonomy level comprises:
querying for the approval feedback signal prior to receiving the approval feedback signal.

5. The method of claim 3 ,
wherein the autonomy level is the third autonomy level indicating “supervised autonomy”, and
wherein performing the action according to the autonomy level comprises:
determining that the AV is being monitored by a human before performing the action.

6. The method of claim 3 , further comprising:
receiving the approval feedback signal.

7. The method of claim 1 , wherein performing the action according to the autonomy level comprises:
determining, based on the autonomy level, whether to request approval from a human for the action before performing the action.

8. The method of claim 1 ,
wherein the autonomy level is an “no autonomy” such that the AV is not allowed to perform autonomous actions, and
wherein performing the action according to the autonomy level comprises:
enabling the AV to be manually controlled by a human.

9. The method of claim 1 ,
wherein the autonomy model comprises a utility model and an autonomy profile,
wherein utility model describes a utility of performing a first action in a first autonomy level with respect to a first environment state given that the AV just operated in second autonomy level, and
wherein the autonomy profile maps respective environment states to respective actions and prescribing constraints on allowed levels of autonomy for particular environment states.

10. The method of claim 1 , further comprising:
updating at least one of an autonomy profile, a feedback profile, or a human transition function in response to the performing the action.

11. An autonomous vehicle (AV), comprising:
a processor configured to execute instructions to:
detect, based on sensor data, an environment state;
select an action based on the environment state;
determine an autonomy level associated with the environment state and the action,
wherein the autonomy level indicates an extent to which the action is to be performed autonomously,
wherein the AV determines respective autonomy levels for actions to be performed,
wherein the autonomy level is selected based at least on an autonomy model and a feedback model, and
wherein the feedback model models feedback types received from a human with respect to the action, how costly each type of feedback is, and how likely is each type of feedback to be received from the human; and

cause the action to be performed according to the autonomy level, wherein to cause the action to be performed according to the autonomy level comprises to:
in response to determining that the autonomy level indicates “verified autonomy” and to determining that the AV is being monitored,
cause the action to be performed; and
in response to receiving an override signal while the action is being performed, cause the action to stop and cause a switch to a manual operation mode of the AV.

12. The AV of claim 11 , wherein the autonomy level is selected from a set comprising a first autonomy level indicating “no autonomy”, a second autonomy level indicating “verified autonomy”, a third autonomy level indicating “supervised autonomy”, and a fourth autonomy level indicating “unsupervised autonomy”.

13. The AV of claim 12 ,
wherein the autonomy level is the second autonomy level indicating “verified autonomy”, and wherein to perform the action according to the autonomy level comprises to:
receive, for the action, an approval feedback signal or a disapproval feedback signal.

14. The AV of claim 13 , wherein to perform the action according to the autonomy level comprises to:
query for the approval feedback signal prior to receiving the approval feedback signal.

15. The AV of claim 13 ,
wherein the autonomy level is the third autonomy level indicating “supervised autonomy”, and
wherein to performing the action according to the autonomy level comprises to:
determine that the AV is being monitored by a human before performing the action.

16. The AV of claim 13 , wherein the processor is further configured to execute instructions to:
receive the approval feedback signal.

17. The AV of claim 11 , wherein to perform the action according to the autonomy level comprises to:
determine, based on the autonomy level, whether to request approval from a human for the action before performing the action.

18. The AV of claim 11 ,
wherein the autonomy level is an “no autonomy” such that the AV is not allowed to perform autonomous actions, and
wherein to perform the action according to the autonomy level comprises to:
enable the AV to be manually controlled by a human.

19. The AV of claim 11 ,
wherein the autonomy model comprises a utility model and an autonomy profile,
wherein utility model describes a utility of performing a first action in a first autonomy level with respect to a first environment state given that the AV just operated in second autonomy level, and
wherein the autonomy profile maps respective environment states to respective actions and prescribing constraints on allowed levels of autonomy for particular environment states.

20. The AV of claim 11 , wherein the processor is further configured to execute instructions to:
update at least one of an autonomy profile, a feedback profile, or a human transition function in response to performing the action.

Assignees

• Nissan North America Inc
• Univ Massachusetts
• Renault Sas

Inventors

• Basich Connor
• Wray Kyle Hollins
• Witwicki Stefan
• Zilberstein Shlomo

Classifications

• B60W60/005

  Handover processes (Handing over between remote control and on-board control or handing over between remote control arrangements G05D1/227) · CPC title

• G08G1/167

  Driving aids for lane monitoring, lane changing, e.g. blind spot detection · CPC title

• G08G1/166 Primary

  for active traffic, e.g. moving vehicles, pedestrians, bikes · CPC title

• G08G1/163

  involving continuous checking · CPC title

• G05D1/0088 Primary

  characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours (using knowledge based models G06N5/00) · CPC title