Extending NanoMap for planning and control

[ViWalkerDev]

Hi there all,

Did some work over the last 7 months extending my last work NanoMap, and integrating it into a multi-agent planning and control framework.

Nanomap Publication:

https://www.mdpi.com/2072-4292/14/21/5463

Planning and Control Publication:

https://www.mdpi.com/2072-4292/15/15/3802

The code is not currently up to date with the planning and control code. It will be updated over the next few weeks after I tidy up the code and will be available here: https://github.com/ViWalkerDev/NanoMap/tree/mdpi-planning-and-control-2023

I have a lot of ideas for improvements and optimizations, as this is more a proof of concept with the some glaring limitations that I want to address.

The work takes the existing nanomap library and extends it to enable planning and control functionality for search/target-finding, and mapping problems. As well as adding a python wrapper and openai gym compatibility so agent policies can be trained using SB3.

This is the video showing some of the results.

https://www.youtube.com/watch?v=FWLTIBuQVEw&t=9s

Will continue working on it, as I have some big ideas for optimizations and improvements to performance, flexibility, and scalability.