Off road assistance
A method for off road driving, the method may include obtaining environment sensed information about an environment of a vehicle of a certain model, by one of more vehicle sensors of the vehicle and while driving over an off road path; detecting, by a machine learning process, an off road driving event; determining, by the machine learning process, a characteristic behavior of vehicles of the certain model when facing the off road driving event; and responding, at least in part by the machine learning process, to the occurrence of the off road driving event.
1. A method for off road driving, the method comprises:
obtaining sensed information pertaining to an environment of an ego vehicle, by one of more vehicle sensors of the vehicle and while driving over an off road path;
detecting, by a machine learning process, an occurrence of an off road driving event; wherein the detecting comprises generating a signature of the sensed information by:
performing a first number of signature generating iterations using a convolutional neural network of fixed connectivity; and
performing additional signature generation iterations that involve deactivating, per each additional signature generation iteration, spanning elements that are irrelevant to the additional signature generation iteration;
determining, by the machine learning process, a characteristic behavior of a vehicle of a model similar to the ego vehicle when facing the off road driving event; and
responding, at least in part by the machine learning process, to the occurrence of the off road driving event; wherein the responding is associated with least one of autonomously driving the ego vehicle or autonomously changing a configuration of the ego vehicle;
wherein the machine learning process is trained by a training process that involves:
obtaining off-road information sensed by vehicles of a model similar to model of the ego vehicle; wherein the off-road information comprises environmental off-road information pertaining to the environment of the vehicles and telematic off-road information indicative of characteristic behaviors of the vehicles;
clustering the off-road environmental information to provide a first group of clusters; and
generating a second group of clusters, based on the first group, by clustering the first group of clusters based on the off-road telematic information, such that different clusters of the second group of clusters are associated with different characteristic behaviors of the vehicles of the model similar to the model of the ego vehicle, and each different characteristic behavior is among a defined number of most frequently occurring behaviors of a different off-road event when facing different off road driving events.
2. The method according to claim 1 further comprising executing the training process.
3. The method according to claim 1 , wherein an off road driving event is indicative of at least one out of a one or more parameter of one or more obstacle, one or more off road path parameter, one or more weather condition, and one or more illumination condition.
4. The method according to claim 3 wherein the one or more off road path parameter comprises at least one out of a slope of an off road path segment that precedes the vehicle, a curvature of the off road path segment, a roughness of the off road path segment, and a type of the off road path segment.
5. The method according to claim 3 wherein the one or more off road path parameter comprises a presence of mud in an off road path segment that precedes the vehicle, and an estimated difficulty level of passing the mud.
6. The method according to claim 1 , wherein the training process is an unsupervised training process.
7. The method according to claim 1 , wherein the obtaining of the off-road sensed information sensed by vehicles of the certain model comprises selecting the off-road sensed information out of a data set of sensed information that also comprises road sensed information.
8. The method according to claim 1 wherein the responding comprises transitioning control of the vehicle from a human driver of the vehicle.
9. The method according to claim 1 wherein the responding further comprises suggesting a driving path to a driver.
10. The method according to claim 1 wherein the responding comprises performing the autonomous driving of the vehicle.
11. The method according to claim 1 wherein the responding further comprises generating an off road driving event alert.
12. The method according to claim 1 wherein the responding comprises autonomously activating a differential lock of the vehicle.
13. The method according to claim 1 wherein the responding comprises autonomously changing the configuration of the vehicle.
14. A non transitory computer readable medium that stores instructions for:
obtaining sensed information pertaining to an environment of an ego vehicle, by one of more vehicle sensors of the vehicle and while driving over an off road path;
detecting, by a machine learning process, an occurrence of an off road driving event; wherein the detecting comprises generating a signature of the sensed information by:
performing a first number of signature generating iterations using a convolutional neural network of fixed connectivity; and
performing additional signature generation iterations that involve deactivating, per each additional signature generation iteration, spanning elements that are irrelevant to the additional signature generation iteration;
determining, by the machine learning process, a characteristic behavior of a vehicle of a model similar to the ego vehicle when facing the off road driving event; and
responding, at least in part by the machine learning process, to the occurrence of the off road driving event; wherein the responding is associated with least one of autonomously driving the ego vehicle or autonomously changing a configuration of the ego vehicle;
wherein the machine learning process is trained by a training process that involves:
obtaining off-road information sensed by vehicles of a model similar to model of the ego vehicle; wherein the off-road information comprises environmental off-road information pertaining to the environment of the vehicles and telematic off-road information indicative of characteristic behaviors of the vehicles;
clustering the off-road environmental information to provide a first group of clusters; and
generating a second group of clusters, based on the first group, by clustering the first group of clusters based on the off-road telematic information, such that different clusters of the second group of clusters are associated with different characteristic behaviors of the vehicles of the model similar to the model of the ego vehicle, and each different characteristic behavior is among a defined number of most frequently occurring behaviors of a different off-road event when facing different off road driving events.
15. The non transitory computer readable medium according to claim 14 further storing instructions for executing the training process.
16. The non transitory computer readable medium according to claim 15 wherein the generating comprises performing one or more clustering iterations until providing the different second clusters are associated with different characteristic behaviors of vehicles of the certain model when facing different off road driving events.
17. The non transitory computer readable medium according to claim 15 wherein an off road driving event is indicative of at least one out of a one or more parameter of one or more obstacle, one or more off road path parameter, one or more weather condition, and one or more illumination condition.
18. The non transitory computer readable medium according to claim 17 wherein the one or more off road path parameter comprises at least one out of a slope of an off road path segment that precedes the vehicle, a curvature of the off road path segment, a roughness of the off road path segment, and a type of the off road path segment.
19. The non transitory computer readable medium according to claim 17 wherein the one or more off road path parameter comprises a presence of mud in an off road path segment that precedes the vehicle, and an estimated difficulty level of passing the mud.