I am a Robotics researcher, as my official job title clearly states. However, my view of what this means – in terms of scientific questions I address – is much broader than most people typically assume. For instance, these days, I spend a lot of time thinking about stochastic control, Markov Decision Processes, reinforcement learning and other related issues such as dynamics abstraction and robust control. On the face of it, these topics seem much broader than the narrow definition of getting robots to do the things that appear in movies and spark the popular imagination. However, as most researchers in this area know, the path to capable robots runs precisely through these sorts of fundamental conceptual issues. This is what attracted me to robotics originally and, I think, this is what makes it a worthwhile area of serious study.
Recently, I came across this short (just a few slides) presentation by Matt Mason, Director of CMU’s Robotics Institute, that makes this point – we need to look beyond the robot to understand the true impact of robotics. Often, people evaluate the field exclusively by the few industrial products that come out of it and completely ignore the much wider scope and impact of the field. The real story is much more interesting!