The robot (Fetch) is going to pick up something on a table in front of him. I added the octomap but the robot keeps colliding with the table, especially the table edge.
I think it is because the generated trajectory points are too close to obstacles and the generated trajectory is too sparse (around 12 trajectory points). Even though all trajectory points are avoiding obstacles but the robot arm still moves through table edge because the line between one trajectory point and the another one are through the table edge. (See the pictures appended)
I am currently using `setPoseTarget`of the `MoveGroup` class (`moveit::planning_interface::MoveGroup`). I tried the `computeCartesianPath` method because I can fine-tune the `double eef_step` parameter to but it won't give me a solution mostly, i.e., no motion plan found.
So my questions are:
1. Is there a way to set the distance to obstacles (or octomap obstacles specifically)?
2. Is there a way to increase the number of trajectory points in `MoveGroup` so `move_group.plan` can return more trajectory points?
I am using Melodic on Ubuntu 18.04 and the binary version of moveit (1.0.1-0bionic.20190320.192509).
Here are some log from terminal:
[ INFO] [1555596197.327901434] [MoveGroupMoveAction::executeMoveCallback_PlanOnly]: Planning request received for MoveGroup action. Forwarding to planning pipeline.
[ INFO] [1555596197.331271450] [ModelBasedPlanningContext::useConfig]: Planner configuration 'arm_with_torso' will use planner 'geometric::RRTConnect'. Additional configuration parameters will be set when the planner is constructed.
[ INFO] [1555596197.332833512] [OMPLPlannerManager::OMPLPlannerManager]: RRTConnect: Starting planning with 1 states already in datastructure
[ INFO] [1555596197.364304111] [OMPLPlannerManager::OMPLPlannerManager]: RRTConnect: Created 6 states (2 start + 4 goal)
[ INFO] [1555596197.364438718] [OMPLPlannerManager::OMPLPlannerManager]: Solution found in 0.032557 seconds
[ INFO] [1555596197.414208802] [OMPLPlannerManager::OMPLPlannerManager]: SimpleSetup: Path simplification took 0.049537 seconds and changed from 5 to 5 states
trajectory point n (on the right side of the table):
trajectory point n+1 (on the back side of the table)
↧