Safe transfer between manned and autonomous driving modes
A method for safe transfer between manned and autonomous driving modes, the method may include detecting, based on first sensed information sensed during a first period, a situation related to an environment of the autonomous vehicle; searching for one or more matching concepts of a group of reference concepts, to which the situation belongs, each reference concept of the group represents a plurality of situations and has a reference concept safety level; wherein for each reference concept of at least a sub-group of the group of reference concepts the safety level of the reference concept is based on a tested success level of at only some of the plurality of scenarios represented by the reference concept; and determining, based on an outcome of the searching, whether the vehicle is capable to safely autonomously drive through the environment.
1. A method for safe transfer between manned and autonomous driving modes, the method comprises:
detecting, by using a machine learning process, a situation related to an environment of an autonomous vehicle, based on first sensed information sensed during a first period using a situation signature generated by limiting, without applying a recognition process to the first sensed information, a processing of the first sensed information according to statistical probabilities of appearances of situation pertaining information within different parts of the first sensed information, such that the situation signature is generated in association with the first sensed information; wherein the limiting, without applying the recognition process to the first sensed information, is (a) also based on an inclination of the autonomous vehicle, and (b) further comprises determining whether to process only a lower part of the first sensed information, wherein the lower part of the first sensed information is defined based on the inclination of the autonomous vehicle and according to an estimation that the autonomous vehicle is located ahead of a T-junction;
searching, based on the generated situation signature, a matching concept that is related to the situation, the matching concept belongs to a group of reference concept structures associated with the first sensed information;
each reference concept structure of the group represents a plurality of situations and is associated with a reference concept structure safety level; wherein each reference concept structure comprises a reference concept structure signature, reference signatures and reference metadata related to the reference signatures;
wherein for each reference concept structure of at least a sub-group of the group of reference concept structures the reference concept structure safety level of the reference concept structure is based on a tested success level of only some of the plurality of scenarios represented by the reference concept structure;
determining, based on an outcome of the searching, whether the autonomous vehicle is capable to safely autonomously drive through the environment; and
outputting, based on the determining, a transition control instruction, wherein the transition control instruction is related to a transition of control of the autonomous vehicle from autonomous driving to human assisted driving of the autonomous vehicle when determining that the autonomous vehicle is not capable to safely autonomously drive through the environment and when the determining whether the autonomous vehicle is capable to safely autonomously drive through the environment occurred while autonomously driving the autonomous vehicle.
2. The method according to claim 1 , wherein the method comprises generating the situation signature, and searching for the one or more matching concept structures by comparing the situation signature to at least one type of signatures out of reference concept structure signatures and reference signatures of the group of reference concept structures.
3. The method according to claim 1 wherein at least some of the reference concept structures differ from each other by complexity of maneuvers required to overcome one or more situations of at least some of the reference concept structures.
4. The method according to claim 1 wherein at least some of the reference concept structures differ from each other by resolution, wherein a resolution of a reference concept structure equals a length of a reference signature that belongs to the reference concept structure, wherein the length of the reference signature that belongs to the reference concept structure is a number of indexes included in the reference signature.
5. The method according to claim 1 , further comprising detecting, in proximity to the first period, one or more objects located within the environment of the autonomous vehicle; and estimating one or more behavior patterns of the one or more objects located within the environment of the autonomous vehicle.
6. The method according to claim 1 , further comprising detecting, in proximity to the first period, one or more objects located within the environment of the autonomous vehicle, wherein the one or more objects are road users and wherein the determining is also responsive to a degree of predictability of movement of road users.
7. A system for safe transfer between manned and autonomous driving modes, the system comprises a processing unit and a memory unit;
wherein the memory unit is configured to store first sensed information sensed during a first period;
wherein the processing unit is configured to:
detect, by using machine a learning process, a situation related to an environment of an autonomous vehicle, based on first sensed information sensed during the first period using a situation signature generated by limiting, without applying a recognition process to the first sensed information, a processing of the first sensed information according to statistical probabilities of appearances of situation pertaining information within different parts of the first sensed information; wherein the limiting, without applying the recognition process to the first sensed information, is (a) also based on an inclination of the autonomous vehicle, and (b) further comprises determining whether to process only a lower part of the first sensed information, wherein the lower part of the first sensed information is defined based on the inclination of the autonomous vehicle and according to an estimation that the autonomous vehicle is located ahead of a T-junction;
search, based on the situation signature, a matching concept that is related to the situation, the matching concept belongs to a group of reference concept structures;
each reference concept structure of the group represents a plurality of situations and is associated with a reference concept structure safety level, for each reference concept structure of at least a sub-group of the group of reference concept structures the reference concept structure safety level is based on a tested success level of only some of the plurality of scenarios represented by the reference concept structure; and
determine, based on an outcome of the searching, whether the autonomous vehicle is capable to safely autonomously drive through the environment; and
outputting, based on the determining, a transition control instruction, wherein the transition control instruction is related to a transition of control of the autonomous vehicle from autonomous driving to human assisted driving of the autonomous vehicle when determining that the autonomous vehicle is not capable to safely autonomously drive through the environment and when the determining of whether the autonomous vehicle is capable to safely autonomously drive through the environment occurred while autonomously driving the autonomous vehicle.
8. The system according to claim 7 , wherein the processing unit is configured to generate the situation signature, and search for the one or more matching concept structures by comparing the situation signature to at least one type signatures of reference concept structure signatures and reference signatures of the group of reference concept structures.
9. The system according to claim 7 wherein at least some of the reference concept structures differ from each other by complexity of maneuvers required to overcome one or more situations of at least some of the reference concept structures.
10. The system according to claim 7 wherein at least some of the reference concept structures differ from each other by resolution, wherein a resolution of a reference concept structure equals a length of a reference signature that belongs to the reference concept structure, wherein the length of the reference signature that belongs to the reference concept structure is a number of indexes included in the reference signature.
11. The system according to claim 7 wherein the processing unit is configured to detect, in proximity to the first period, one or more objects located within the environment of the autonomous vehicle.
12. The system according to claim 11 wherein the processing unit is configured to estimate one or more behavior patterns of the one or more objects located within the environment of the autonomous vehicle.
13. The system according to claim 12 wherein the one or more objects are one or more road users and wherein the determining comprises estimating one or more behavior patterns of the one or more road users.
14. A non-transitory computer readable medium that stores instructions for:
detecting, by using a machine learning process, a situation related to an environment of an autonomous vehicle, based on first sensed information sensed during a first period using a situation signature generated by limiting, without applying a recognition process to the first sensed information, a processing of the first sensed information according to statistical probabilities of appearances of situation pertaining information within different parts of the first sensed information; wherein the limiting, without applying the recognition process to the first sensed information, is (a) also based on an inclination of the autonomous vehicle, and (b) further comprises determining whether to process only a lower part of the first sensed information, wherein the lower part of the first sensed information is defined based on the inclination of the autonomous vehicle and according to an estimation that the autonomous vehicle is located ahead of a T-junction;
searching, based on the situation signature, a matching concept that is related to the situation, the matching concept belongs to a group of reference concept structures;
each reference concept structure of the group represents a plurality of situations and is associated with a reference concept structure safety level, for each reference concept structure of at least a sub-group of the group of reference concept structures the reference concept structure safety level is based on success levels of multiple attempts to autonomously drive one or more autonomous vehicles at only some of the plurality of scenarios represented by the reference concept structure; wherein each reference concept structure comprises a reference concept structure signature, reference signatures and reference metadata related to the reference signatures;
determining, based on an outcome of the searching, whether the autonomous vehicle is capable to safely autonomously drive through the environment; and
outputting, based on the determining, a transition control instruction, wherein the transition control instruction is related to a transition of control of the autonomous vehicle from autonomous driving to human assisted driving of the autonomous vehicle when determining that the autonomous vehicle is not capable to safely autonomously drive through the environment and when the determining whether the autonomous vehicle is capable to safely autonomously drive through the environment occurred while autonomously driving the autonomous vehicle.
15. The non-transitory computer readable medium according to claim 14 , wherein the non-transitory computer readable medium further stores instructions for generating the situation signature; and searching for the one or more matching concept structures by comparing the situation signature to at least one type of signatures of reference concept structure signatures and reference signatures of the group of reference concept structures.
16. The non-transitory computer readable medium according to claim 14 wherein at least some of the reference concept structures differ from each other by complexity of maneuvers required to overcome one or more situations of at least some of the reference concept structures.
17. The non-transitory computer readable medium according to claim 14 wherein at least some of the reference concept structures differ from each other by resolution, wherein a resolution of a reference concept structure equals a length of a reference signature that belongs to the reference concept structure, wherein the length of the reference signature that belongs to the reference concept structure is a number of indexes included in the reference signature.
18. The non-transitory computer readable medium according to claim 14 , further storing instructions for detecting, in proximity to the first period, one or more objects located within the environment of the autonomous vehicle.
19. The non-transitory computer readable medium according to claim 18 wherein the one or more objects are one or more road users and wherein the non-transitory computer readable medium stores instructions for estimating one or more behavior patterns of the one or more road users.
20. The non-transitory computer readable medium according to claim 14 wherein the situation comprises a weather condition and a potential risk, wherein the potential risk is based on one or more risk factors that comprises illumination of the environment and traffic load.
21. The non-transitory computer readable medium according to claim 14 , wherein the autonomous driving further comprises driving the autonomous vehicle through the environment.