Robotics research combines some of the best challenges from optimization, differential equations, machine learning, and high-performance computing. Robots are complex dynamical systems, and the fundamental questions of how to coordinate and control all but the simplest of robots are still open. It’s up to us as scientists and engineers and makers to bring about the best robotic future we can, and Julia will be part of that future. In this talk, I’ll discuss how we’re replacing Python, C++, and MATLAB with a new set of tools in Julia to simulate, visualize, and control robots. I’ll also introduce the JuliaRobotics organization, which has been founded to coordinate and promote robotics software development in Julia. In particular, I’ll discuss our work at MIT towards a major goal: controlling a 350 pound hydraulic walking humanoid robot in Julia. In the process, I’ll talk about the custom tools that we’ve built in JuliaRobotics as well as the ways we’re integrating with the excellent tools from JuliaDiffEq, JuliaOpt, FluxML, JuliaGeometry, JuliaGizmos, and more. You’ll hear about:
Robin is a PhD student at MIT who should probably be working on his thesis instead of writing JuliaCon talk proposals. His early PhD work involved planning and control of the Atlas humanoid robot at the DARPA Robotics Challenge, and he is now interested in helping walking robots safely move through the world without falling. He’s a recovering former MATLAB user, and he’s been developing robotics tools in Julia for the past few years. You can see some of his work on github some of his silly projects on his blog and some of his thoughts on Julia for robotics at Julia Computing’s case studies.