Published: Dec 18, 2022 by Liz Metzger
(Wiring, Manufacturing, CAD)
The STALKER project was a first pass at making a five legged, walking battlebot. I worked on this project during an internship and got to be a part of it from the earliest stages where STALKER was just an idea and some servos laid out rather haphazardly on a table. The design of this robot is rather reminiscent of Axis, one of the other projects on my portfolio, but with more degrees of freedom and a lot more power. I was responsible for getting all of the 3D pieces off of the printing bed to being assembled, the wiring, and soldering.
The first step was to get a labeling scheme for the servos that made sense so that we would be able to effectively talk about what was broken where, what still needed to get put together, and what was still missing parts. We decided that each leg would be numbered 1-5 and that the servos in each leg would also be numbered 1-4 so the numbering scheme for the first leg was 110, 111, 112, and 113. The second leg was numbered 120, 121, 122, and 123 and so on and so forth for each leg. Once we had our numbering scheme down all of the servos had to be labeled and commissioned.
After making a plan we started to print, laser cut, and assemble STALKER. This process took a long time as we realized that parts that were designed needed new cutouts to pass cables through or didn’t have the correct clearances. It was also challenging to work with because as we began to attach servos to each other the whole assembly started to get really heavy and hard to handle. As we started to get STALKER put together we had a better idea of what length cables we needed where to connect all of the servos to the central board and to power. I made custom wiring for each leg by cutting, stripping, and crimping wires to chain the servos together.
Around the time that STALKER was looking like it above we started to think of a control system. Our first approach was to try to use geometry to generate the IK of the robot in a software called MAX. Since the servos were not all moving in the same plane and were offset from each other the geometry was super complicated to figure out and required us to make assumptions that simplified it significantly. We did not think that this would be an acceptable control system and started working out how to control the robot using rotation matrices. Although by the end of my internship we didn’t have a fleshed out control system, we did have an entirely assembled robot and were able to stand it up!
Since my internship the control system has been fully fleshed out and STALKER has been demoed at multiple events including battlebots.