Hierarchical planning for autonomous machinery on various tasks
Autonomous excavator has developed rapidly in recent years because of a shortage of labor and hazardous working environments for operating excavators. Presented herein are embodiments of a novel hierarchical planning system for autonomous machines, such as excavators. In one or more embodiments, the overall planning system comprises a high-level task planner for task division and base movement planning, and general sub-task planners with motion primitives, which include both arm and base movement in the case of an excavator. Using embodiments of the system architecture, experiments were performed for the trench and pile removal tasks in the real world and for the large-scale material loading tasks in a simulation environment. The results show that the system architecture embodiments and planner method embodiments generate effective task and motion plans that perform well in autonomous excavation.
1. A computer-implemented method comprising:
receiving, from a user, a task for an excavator and associated task parameter values;
given the task, the associated task parameter values, at least some of perception-related data received from a perception module, using a high-level task planner module comprising one or more rules-based models to divide the task into one or more base move sub-tasks and one or more material removal sub-tasks, wherein the one or more base move sub-tasks comprise a sequence of route points and the one or more material removal sub-tasks comprise a sequence of sub-task regions and a goal for each sub-task region;
given the sequence of route points and states information wherein at least some of the states information is obtained from the perception module, using a base move sub-task planner that comprises one or more models to generate a set of one or more waypoints between consecutive pairs of route points to guide the excavator between route points;
given the sequence of sub-task regions, associated goals, and states information wherein at least some of the states information is obtained from the perception module, using a material removal sub-task planner comprising one or more planner models to determine motion primitives and related parameter values for each sub-task region to accomplish the goal for that sub-task region;
using the motion primitives, related parameters, excavator parameters, and perception module data as inputs to a joint move primitive module to generate joint trajectories and using waypoints, excavator parameters, and perception module data as inputs to a base move primitive module to generate base trajectories; and
providing the joint trajectories and base trajectories to joint controller module(s) and base controller module(s), respectively, to generate controller commands to control devices of the excavator.
2. The computer-implemented method of claim 1 wherein the perception module comprises or receives data from one or more sensors comprising:
a camera for image data;
a LiDAR sensor for generating point cloud data;
a real-time kinematic sensor for localization; and
one or more inclination sensors.
3. The computer-implemented method of claim 1 further comprising determining a local task region using a rectangular definition or sector-based definition and a set of parameter values comprising area center swing angle, area angle range or width, near-end distance, and remote-end distance.
4. The computer-implemented method of claim 3 wherein the sub-task regions are determined by the high-level task planner by performing steps comprising:
given the local task region for the excavator and an area over which the task is to be performed, determining a number of local task regions that at least covers the area.
5. The computer-implemented method of claim 1 wherein the step of given the sequence of sub-task regions, associated goals, and state information obtained from the perception module, using a material removal sub-task planner comprising one or more models to determine motion primitives and related parameter values for each sub-task region to accomplish the goal for that sub-task region comprises:
for a sub-task region, implementing a set of one or more action planners according to a material removal sub-task planner model associated with the sub-task at the sub-task region until the sub-task goal is achieved; and
responsive to the sub-task goal for the sub-task region being achieved and responsive to the sub-task region not being the last sub-task region in the sequence of sub-task regions, processing a next sub-task region in the sequence of sub-task regions.
6. The computer-implemented method of claim 1 wherein the material removal sub-task planner comprises:
a point-of-attack (POA) planner;
a dig motion planner; and
a dump planner.
7. The computer-implemented method of claim 6 wherein the dig motion planner utilizes parameters comprising:
x b , y b , z b , which represents a point-of-attack (POA) location in base frame;
p d , which represents penetration depth;
p l , which represents penetration length;
d l , which represents drag length, and
δ, which represents entire dig motion tilt angle with respect to horizon plane.
8. The computer-implemented method of claim 1 wherein:
the states information for the base move sub-task planner comprises excavator joint states, excavator base pose, and one or more environment states of a working area of the excavator; and
the states information for the base move sub-task planner comprises excavator base pose and one or more environment states of a working area of the excavator.
9. The computer-implemented method of claim 1 wherein an end condition of a sub-task goal may be defined according to different levels of strictness requirement.
10. A system comprising:
one or more processors; and
a non-transitory computer-readable medium or media comprising one or more sets of instructions which, when executed by at least one of the one or more processors, causes steps to be performed comprising:
receiving, from a user, a task for an excavator and associated task parameter values;
given the task, the associated task parameter values, at least some of perception-related data received from a perception module, using a high-level task planner module comprising one or more rules-based models to divide the task into one or more base move sub-tasks and one or more material removal sub-tasks, wherein the one or more base move sub-tasks comprise a sequence of route points and the one or more material removal sub-tasks comprise a sequence of sub-task regions and a goal for each sub-task region;
given the sequence of route points and states information wherein at least some of the states information is obtained from the perception module, using a base move sub-task planner that comprises one or more models to generate a set of one or more waypoints between consecutive pairs of route points to guide the excavator between route points;
given the sequence of sub-task regions, associated goals, and states information wherein at least some of the states information is obtained from the perception module, using a material removal sub-task planner comprising one or more planner models to determine motion primitives and related parameter values for each sub-task region to accomplish the goal for that sub-task region;
using the motion primitives, related parameters, excavator parameters, and perception module data as inputs to a joint move primitive module to generate joint trajectories and using waypoints, excavator parameters, and perception module data as inputs to a base move primitive module to generate base trajectories; and
providing the joint trajectories and base trajectories to joint controller module(s) and base controller module(s), respectively, to generate controller commands to control devices of the excavator.
11. The system of claim 10 wherein the perception module comprises or receives data from one or more sensors comprising:
a camera for image data;
a LiDAR sensor for generating point cloud data;
a real-time kinematic sensor for localization; and
one or more inclination sensors.
12. The system of claim 10 wherein the non-transitory computer-readable medium or media further comprises one or more sets of instructions which, when executed by at least one of the one or more processors, causes steps to be performed comprising:
determining a local task region using a rectangular definition or sector-based definition and a set of parameter values comprising area center swing angle, area angle range or width, near-end distance, and remote-end distance.
13. The system of claim 10 wherein the step of given the sequence of sub-task regions, associated goals, and state information obtained from the perception module, using a material removal sub-task planner comprising one or more models to determine motion primitives and related parameter values for each sub-task region to accomplish the goal for that sub-task region comprises:
for a sub-task region, implementing a set of one or more action planners according to a material removal sub-task planner model associated with the sub-task at the sub-task region until the sub-task goal is achieved; and
responsive to the sub-task goal for the sub-task region being achieved and responsive to the sub-task region not being the last sub-task region in the sequence of sub-task regions, processing a next sub-task region in the sequence of sub-task regions.
14. The system of claim 10 wherein the material removal sub-task planner comprises:
a point-of-attack (POA) planner;
a dig motion planner; and
a dump planner.
15. The system of claim 14 wherein the dig motion planner utilizes parameters comprising:
x b , y b , z b , which represents a point-of-attack (POA) location in base frame;
p d , which represents penetration depth;
p l , which represents penetration length;
d l , which represents drag length, and
δ, which represents entire dig motion tilt angle with respect to horizon plane.
16. The system of claim 10 wherein:
the states information for the base move sub-task planner comprises excavator joint states, excavator base pose, and one or more environment states of a working area of the excavator; and
the states information for the base move sub-task planner comprises excavator base pose and one or more environment states of a working area of the excavator.
17. A non-transitory computer-readable medium or media comprising one or more sequences of instructions which, when executed by at least one processor, causes steps to be performed comprising:
receiving, from a user, a task for an excavator and associated task parameter values;
given the task, the associated task parameter values, at least some of perception-related data received from a perception module, using a high-level task planner module comprising one or more rules-based models to divide the task into one or more base move sub-tasks and one or more material removal sub-tasks, wherein the one or more base move sub-tasks comprise a sequence of route points and the one or more material removal sub-tasks comprise a sequence of sub-task regions and a goal for each sub-task region;
given the sequence of route points and states information wherein at least some of the states information is obtained from the perception module, using a base move sub-task planner that comprises one or more models to generate a set of one or more waypoints between consecutive pairs of route points to guide the excavator between route points;
given the sequence of sub-task regions, associated goals, and states information wherein at least some of the states information is obtained from the perception module, using a material removal sub-task planner comprising one or more planner models to determine motion primitives and related parameter values for each sub-task region to accomplish the goal for that sub-task region;
using the motion primitives, related parameters, excavator parameters, and perception module data as inputs to a joint move primitive module to generate joint trajectories and using waypoints, excavator parameters, and perception module data as inputs to a base move primitive module to generate base trajectories; and
providing the joint trajectories and base trajectories to joint controller module(s) and base controller module(s), respectively, to generate controller commands to control devices of the excavator.
18. The non-transitory computer-readable medium or media of claim 17 wherein the step of given the sequence of sub-task regions, associated goals, and state information obtained from the perception module, using a material removal sub-task planner comprising one or more models to determine motion primitives and related parameter values for each sub-task region to accomplish the goal for that sub-task region comprises:
for a sub-task region, implementing a set of one or more action planners according to a material removal sub-task planner model associated with the sub-task at the sub-task region until the sub-task goal is achieved; and
responsive to the sub-task goal for the sub-task region being achieved and responsive to the sub-task region not being the last sub-task region in the sequence of sub-task regions, processing a next sub-task region in the sequence of sub-task regions.
19. The non-transitory computer-readable medium or media of claim 17 wherein the material removal sub-task planner comprises:
a point-of-attack (POA) planner;
a dig motion planner; and
a dump planner.
20. The non-transitory computer-readable medium or media of claim 17 wherein:
the states information for the base move sub-task planner comprises excavator joint states, excavator base pose, and one or more environment states of a working area of the excavator; and
the states information for the base move sub-task planner comprises excavator base pose and one or more environment states of a working area of the excavator.