Determining control operations for an autonomous vehicle
Determining control operations for an autonomous vehicle may include receiving, by an operational model, an operational model input based on camera data from one or more cameras of an automated vehicle; determining, by the operational model, based on the input, a current environmental state and a predicted environmental state; providing, to a rules module, a differential between the current environmental state and a previously predicted environmental state; and determining, based on the rules module, one or more control operations for the automated vehicle.
1. A method for determining control operations for an autonomous vehicle, the method comprising:
in response to an autonomous vehicle entering an autonomous driving mode, repeatedly performing at a predefined interval:
receiving, by an operational model, an operational model input comprising video data from one or more cameras of the autonomous vehicle and one or more motion vectors associated with the video data;
determining, by the operational model, based on the operational model input, a current environmental state and a predicted environmental state corresponding to a future time;
determining, by a rules module and based on a differential between the current environmental state and a previously predicted environmental state corresponding to a same time as the current environmental state, one or more control operations for the autonomous vehicle for reducing the differential, wherein determining the one or more control operations comprises:
providing the differential and the one or more motion vectors as inputs to the rules module;
determining whether the one or more motion vectors satisfy one or more conditions;
wherein when the one or more motion vectors satisfy the one or more conditions, determining, the rules module, the one or more control operations independent of the differential;
wherein when the one or more motion vectors do not satisfy the one or more conditions, determining, by the rules module, the one or more control operations based on the differential; and
executing the one or more control operations by the autonomous vehicle.
2. The method of claim 1 , further comprising normalizing the video data.
3. The method of claim 1 , further comprising:
generating, by a motion estimation module, based on the video data, the one or more motion vectors; and
wherein the rules module is configured to accept, from the motion estimation model, an input associated with the one or more motion vectors.
4. The method of claim 1 , further comprising storing the predicted environmental state in a cache.
5. The method of claim 4 , further comprising loading the previously predicted environmental state from the cache.
6. An autonomous vehicle for determining control operations for an autonomous vehicle, comprising:
an apparatus configured to perform steps comprising:
in response to the autonomous vehicle entering an autonomous driving mode, repeatedly performing at a predefined interval:
receiving, by an operational model, an operational model input comprising video data from one or more cameras of the autonomous vehicle and one or more motion vectors associated with the video data;
determining, by the operational model, based on the operational model input, a current environmental state and a predicted environmental state corresponding to a future time;
determining, by a rules module and based on a differential between the current environmental state and a previously predicted environmental state corresponding to a same time as the current environmental state, one or more control operations for the autonomous vehicle for reducing the differential, wherein determining the one or more control operations comprises:
providing the differential and the one or more motion vectors as inputs to the rules module;
determining whether the one or more motion vectors satisfy one or more conditions;
wherein when the one or more motion vectors satisfy the one or more conditions, determining, the rules module, the one or more control operations independent of the differential;
wherein when the one or more motion vectors do not satisfy the one or more conditions, determining, by the rules module, the one or more control operations based on the differential; and
executing the one or more control operations by the autonomous vehicle.
7. The autonomous vehicle of claim 6 , wherein the steps further comprise:
generating, by a motion estimation module, based on the video data, the one or more motion vectors; and
wherein the rules module is configured to accept, from the motion estimation model, an input associated with the one or more motion vectors.
8. The autonomous vehicle of claim 6 , wherein the steps further comprise normalizing the video data.
9. The autonomous vehicle of claim 6 , wherein the steps further comprise:
storing the predicted environmental state in a cache; and
loading the previously predicted environmental state from the cache.
10. An apparatus for determining control operations for an autonomous vehicle, the apparatus configured to perform steps comprising:
in response to an autonomous vehicle entering an autonomous driving mode, repeatedly performing at a predefined interval:
receiving, by an operational model, an operational model input comprising video data from one or more cameras of the autonomous vehicle and one or more motion vectors associated with the video data;
determining, by the operational model, based on the operational model input, a current environmental state and a predicted environmental state corresponding to a future time;
determining, by a rules module and based on a differential between the current environmental state and a previously predicted environmental state corresponding to a same time as the current environmental state, one or more control operations for the autonomous vehicle for reducing the differential, wherein determining the one or more control operations comprises:
providing the differential and the one or more motion vectors as inputs to the rules module;
determining whether the one or more motion vectors satisfy one or more conditions;
wherein when the one or more motion vectors satisfy the one or more conditions, determining, the rules module, the one or more control operations independent of the differential;
wherein when the one or more motion vectors do not satisfy the one or more conditions, determining, by the rules module, the one or more control operations based on the differential; and
executing the one or more control operations by the autonomous vehicle.
11. The apparatus of claim 10 , wherein the steps further comprise:
generating, by a motion estimation module, based on the video data, the one or more motion vectors; and
wherein the rules module is configured to accept, from the motion estimation model, an input associated with the one or more motion vectors.
12. The apparatus of claim 10 , wherein the steps further comprise normalizing the video data.
13. The apparatus of claim 10 , wherein the steps further comprise:
storing the predicted environmental state in a cache; and
loading the previously predicted environmental state from the cache.
14. A computer program product disposed upon a non-transitory computer readable medium, the computer program product comprising computer program instructions for determining control operations for an autonomous vehicle system modality switching in an autonomous vehicle that, when executed, cause a computer system of the autonomous vehicle to carry out the steps of:
in response to the autonomous vehicle entering an autonomous driving mode, repeatedly performing at a predefined interval:
receiving, by an operational model, an operational model input comprising video data from one or more cameras of the autonomous vehicle and one or more motion vectors associated with the video data;
determining, by the operational model, based on the operational model input, a current environmental state and a predicted environmental state corresponding to a future time;
determining, by a rules module and based on a differential between the current environmental state and a previously predicted environmental state corresponding to a same time as the current environmental state, one or more control operations for the autonomous vehicle for reducing the differential, wherein determining the one or more control operations comprises:
providing the differential and the one or more motion vectors as inputs to the rules module;
determining whether the one or more motion vectors satisfy one or more conditions;
wherein when the one or more motion vectors satisfy the one or more conditions, determining, the rules module, the one or more control operations independent of the differential;
wherein when the one or more motion vectors do not satisfy the one or more conditions, determining, by the rules module, the one or more control operations based on the differential; and
executing the one or more control operations by the autonomous vehicle.