Weekly Meeting 06/05/2024

[SimoneMic]

Chair:

• @SimoneMic

Attendees:

• @steb6
• @vigisushrutha23
• @PasMarra
• @andrearosasco

Absent:

---

Agenda:

Note: Last week's items should become this week's agenda.

Item	Tag	Responsible	Comment
Calibration of Realsense TODO		@SimoneMic @PasMarra @andrearosasco
Finish the simpler version of the MPC controller for obstacle avoidance	https://github.com/Woolfrey/paper_fast_mpc/issues/3	@vigisushrutha23	Tuned planner parameters, introduce a orientation weight to the optimizer by john, working on implementing it on the code

Additional Discussion Points:

Anything not captured in the agenda.

Discussion Point	Raised By	Tag	Outcome

---

Next Meeting:

Note: These will become the agenda for the next meeting.

Item	Tag	Responsible
Action item.	#Issue	@handle
Superimpose hands to robot ones		@steb6
Added error functions and Lapunyof based controller for the MPC. This is done for fixing the orientation optimisation. Working on writing the Lapunyov function in QP problem form		@Woolfrey @vigisushrutha23
Reformulated the oculus problem: While initialising the setup, supposing both looking ahead, we collect the info about the rotation between the local reference frames		@steb6 @PasMarra
Tested AMI teleoperation: fine-grained manipulation is not possible, 2D perception only and no depth. We will implement our oun solution with the depth from the camera to exploit the 3D world		@steb6
Simulation Slowdown improved by removing some mutex locks, will look further what could be also tempering the performance		@SimoneMic @vigisushrutha23
(MPC controller) Working with John on the planner side to tune the undershooting of the planner		@vigisushrutha23

Conclusions:

A summary of the meeting.

[S-Dafarra]

Tested AMI teleoperation: fine-grained manipulation is not possible, 2D perception only and no depth. We will implement our oun solution with the depth from the camera to exploit the 3D world

Please let me know what you come up with. I would be very interested in integrating it!