Closed gabeabrams closed 9 years ago
Sure :) sounds good. I'll implement that!
On Sat, Sep 19, 2015 at 11:20 AM, sarahangle notifications@github.com wrote:
Note: fallback has to allow the second arm to drop its block and pull away as the first action.
Also, can the one arm we always use and the arm that acts first be the right one?
— Reply to this email directly or view it on GitHub https://github.com/gabeabrams/robotmanipulation/issues/22#issuecomment-141679226 .
Actually, fallback will have to be given some information in that case. move_planner does not take state of gripper into account.
On Sat, Sep 19, 2015 at 11:23 AM, Gabriel Abrams gabeabrams@gmail.com wrote:
Sure :) sounds good. I'll implement that!
On Sat, Sep 19, 2015 at 11:20 AM, sarahangle notifications@github.com wrote:
Note: fallback has to allow the second arm to drop its block and pull away as the first action.
Also, can the one arm we always use and the arm that acts first be the right one?
— Reply to this email directly or view it on GitHub https://github.com/gabeabrams/robotmanipulation/issues/22#issuecomment-141679226 .
What info does it need?
Also, can we change the table areas to -? Like 0 is rand, -1 is randleft, -2 is randright, -3 is home?
Better yet, can we have no target type? Just BlockIDs with >0 being blocks and <1 being locations?
integrated move_planner and controller, now just waiting on action_handler function so that move_planner outputs a list of service calls!
I am using blockID = 1 for open/close gripper. My functions should be good to integrate. I can't test them without integration though (I can't run the code because it has ROS dependencies)
Note: fallback has to allow the second arm to drop its block and pull away as the first action.
Also, can the one arm we always use and the arm that acts first be the right one?