Hello everyone. I’m a student studying scientific computing and robotics, and some of my recent research has been into software tooling and methods that might help improve the development process for robotics software. I’m still in very early stages, so I’m definitely not promising to revolutionize the field or anything.
I’m looking for a robot design that could act as a test platform for embedded software. I’m looking into embedded real-time control loops, but most of the applications I’ve studied so far are concerned with walking robots. Those are way outside of my budget, and I lack the mechanical expertise to try and design a cheaper one, so I have to find a different set of problems to tackle. In other words, I need a robot that:
- Requires real-time control to function properly.
- Is sufficiently complex that it will fail if my software is no good.
- Has some amount of flexibility in its function, to allow me to test my methods against a wider set of goals.
Right now I’m leaning lightly towards a quadcopter of some sort, but that might be a bit too prone to damage in the case of failure. I’d prefer not to use a basic wheeled/tracked bot, since it’d likely be too simple to ever provide compelling evidence that a given real-time controller is doing anything impressive. If possible, I’d like to stick to open-source designs, particularly those that leave some room for expansion or modification in case of future experiments. I have access to my university’s fablab, so I can make use of materials like laser-cut acrylic if I have to.
Thank you for any advice you may be able to offer. If you need any clarifications on my needs, please feel free to ask.