Battery charging

[LukeAnger1]

Getting a script to charge the battery, created this for reference, will start using this more than slack because I like it better.

[Cormac0]

I created a new branch for working on this - battery_logic. It's split off from integration_final1.

[LukeAnger1]

[LukeAnger1]

Note Mar 24, 2025.pdf

[LukeAnger1]

This is kinda a crude outline of what I did. I was thinking about trying to find these optimal turns and driving distances but looking through cameras more to see if there is constantly an apriltag visible. Thoughts on this? I am unsure.

Also how do you want to do the coding @Cormac0? I would assume either push same branch or just update each others branches as we go along.

[Cormac0]

This is kinda a crude outline of what I did. I was thinking about trying to find these optimal turns and driving distances but looking through cameras more to see if there is constantly an apriltag visible. Thoughts on this? I am unsure.

Also how do you want to do the coding @Cormac0? I would assume either push same branch or just update each others branches as we go along.

Using the same branch probably makes the most sense, if you don't have any strong opinion? I put a boiler plate "charging_navigator" script in the navigation folder that I think could be a good framework for interfacing with the agent. No actual code to allow for charging yet, though.

[Cormac0]

Note Mar 24, 2025.pdf

This looks like a solid plan to me. I think a big benefit is that while we're translating we'll theoretically be able to keep an eye on the AprilTags for localization. It'll only be when we rotate that we'll need to rely on something like the IMU - but that's actually pretty good at just tracking the orientation over time.

[LukeAnger1]

Ok that sounds good. Def not my specialty with that imu/apriltag stuff.

I am thinking about saving all the camera images at these points and the pytransform relative to the global frame at each one. Is there anything else that we could potentially need?

[Cormac0]

Nothing off the top of my head. I wonder if a way to interface with your existing navigator code could be good. That way we could give a target point for it to move to?

[Cormac0]

Took a first pass at a routine that assumes it starts looking straight at the antenna. In theory it should drive straight, rotate, and then flick up the radiator cover. The agent runs in the simulator, but I haven't fully tested it since the --xy argument to run_agent.py wasn't orienting the rover like I'd expected (woops).

If nothing else, I think the agent and the charging_navigator scripts should make for a good framework for how we can interface with the charging routine.

[LukeAnger1]

[LukeAnger1]