Link

Term Reflections (2021 05 26)

As is tradition, the weeks pass with more momentum than the projects which are trying to keep up.

I would say the biggest progress has been made inside of OSAP, which I document here - there I resolved software and hardware modules under one routing and flow control scheme. I like the system, and the bus I developed in order to bring some realtime controllers to life within that system is also alright, I’ve documented that here. The idea with OSAP (open systems assembly protocol) is to develop a software layer inter-networking / interoperability scheme that allows heterogeneous compute / hardware resources to be easily added to a system’s control scheme without enforcing any hardware specification: network links are virtualized so we can add whatever link we would like: TCP/IP based websockets, bluetooth links, usb serial, or whatever low level bus / hardware (UART, SPI, I2C), or i.e. raw ethernet frames. The question is about how open hardware development networks can emerge in a world that is typically dominated by proprietary network control systems and control busses. Open source hardware developers currently have a hard time making atomic contributions like motor controllers, sensors, etc - because (we suppose) integrating specialized devices into larger control networks is so cumbersome: we often have to re-engineer or modify someone else’s specialized work into the confines of our systems’ “main” controller.

I’ve also made some progress on my ‘going’ machine platform clank and I documented that progress here.

Finally, a toolchanger is underway here - this allows multiple end-effectors to share motion platforms - namely clank, again. The toolchanger is very easy to build, and stiff enough (untested claim) to use in a lightweight milling machine.

In all, I wasn’t able to make as much progress across machine designs / component designs than I would have liked this term: I was also taking a feedback controls course that really challenged my maths background. However! I can do state space control (or can at least start to practice with it) and I will be doing a crash course / documentation run at that sometime over the next month.