Controls: State estimation & localization

[stanley-wei]

Look into common strategies for implementing state estimation & localization in AUVs. This can be both AUVs in the wild, or AUVs made by other RoboSub teams specifically. If there are multiple strategies, evaluate relative pros and cons.

Later in the process, once a few potential strategies have been identified, look into implementations, particularly implementations in ROS2.

I made a document with a few initial resources; please add any findings, links, etc. as you go.

Some potential places you might want to look:

• Past technical reports from other RoboSub teams
• Textbooks and/or literature reviews on state estimation & localization in AUVs
• GitHub repositories relating to state estimation & localization in AUVs.
• Any ROS (preferably ROS2) resources relating to state estimation & localization in AUVs.
• Any general resources relating to state estimation & localization (not necessarily in AUVs).

[stanley-wei]

For state estimation: we'll most likely end up using some kind of Kalman filter, namely the Extended Kalman Filter; that being said, there may be alternatives out there that we aren't aware of.

For localization: although we'll likely be primarily relying on the IMU, one possibility could be using visual odometry as an additional source of pose estimates.