“Alex, what are robots with proprioception?”

This article reminds me of a chat about proprioception I had with teammate/clubmate Greg Loomis over the winter. Ok, so our talk in no way involved robots, but it stuck with me nonetheless. I think it’s that they’re using proprioception’s local sensor/local neural loop model to avoid the complexity of integrated analytical system. Greg, what do you think about Runbot?

There are 2 comments

  1. Loomdog

    Amazing! And yes the idea of using peripheral control loops is comparable to proprioception and the "central pattern generator" that humans use to walk. (both are located at levels below the brian… (CPG in the brainstem, and proprio. in the peripheral joint receptors signalling to the spinal cord) Trying to control everything with the brain itself is just too difficuult….and makes you walk "roboticly" You see this in amputees when they first begin to walk with a prosthesis. They try to think through ever single movement…. Then they get it, relax, and begin to "walk naturally" …Awesome.

  2. Trail Goat

    Loomdog,I thought you'd appreciate the article. This morning the article got me thinking about another conversation we had in Hawaii. (I now realize that's where we talked of proprioception.) We brainstormed some ideas for making running and running related activities as career. Well this article got me thinking what if rather than working to apply our ideas to running, we applied the ideas of running to other areas. You know, think out side the box. Bring seemingly completely unrelated subjects together with the ridiculously effective and efficient systems of the human body? In the environmental law journals that I occasionally read, I've read of the concept of biological systems engineering (or something like that), where manufacturing process designs mimic those in nature. The mimicry is not necessarily a step-by-step mimicry, but recycling of unused inputs and seemingly extraneous outputs (whether benign or "waste"). What got my mind racing this morning was not the idea of actually copying our physical or chemical reactions, but stepping back a step and observing our systems more broadly and then seeing if the conceptual framework of those systems has potential applicability to human interaction, production, distribution, or whatever. It's genius that the runbot scientists stepped back and said, the whole central CPU for a walking robot isn't working out so well.. let's see what the body does. (I'm assuming that this was their thought process.) The idea that first popped into my mind was distribution systems. Isn't it amazing is it the way our bodies produce, store, redistribute, and then utilize glycogen…. and how applicable might that be to supply chain management. What if Walmart or another company that incrementally tweaks its supply chain management took a giant step back and look at the system more broadly. What is there are lessons to be learned about when to signal shipping? (how low to let warehouse stock get?) Who should request resupply? (the cell/retail store or the glycogen store/warehouse) Who decides whether to ship more product to the store and to the warehouse? (can the warehouse or CEO object and, if so, how?) I think you get the idea. Ok, so I'm just going crazy.BUT(!), while new products and services come along and are good and all, new systems and organization have the potential to revolutionize all products and services. Isn't this the information age? Isn't it all about how you collect, analyze, and apply data? The body's is pretty good at that as we both know through ultrarunning.

Post Your Thoughts