Sethu Vijayakumar is the Professor of Robotics and the Director of the Institute for Perception, Action and Behavior (IPAB) at the School of Informatics at the University of Edinburgh. He also holds additional appointments as an Adjunct Faculty at the University of Southern California (USC), Los Angeles and as a Visiting Research Scientist at the RIKEN Brain Science Institute, Japan.
Sethu joined the University of Edinburgh in 2003, after a PhD from Tokyo (1998) followed by an adjunct faculty position at USC, Los Angeles and was promoted to his Chair in Robotics in August 2010. His research interest spans the fields of statistical machine learning, robotics, sensorimotor control, and computational neuroscience. Sethu is the author of over 120 highly cited publications in these fields and has pioneered the use of large scale machine learning techniques for real-time, online adaptive control of anthropomorphic robotic systems. He is the winner of the IEEE Vincent Bendix award, the Japanese Monbusho fellowship, the NEC C&C Student award besides serving on numerous EU and DFG grant review panels and program committees of leading international machine learning and robotics conferences. Sethu had led several national and international research projects funded through research councils such as EPRSC, the European Union, Royal Society and RAEng as well as the industrial partners such as HONDA, Touch Bionics and Microsoft. Since August 2007, he holds the prestigious Senior Research Fellowship of the Royal Academy of Engineering, co-funded by Microsoft Research.
2012 is the year of the London Olympics and, appropriately, this talk is about making robots run faster, jump higher and throw further while being as versatile, robust and adaptive as humans. We are very adept at performing fast, complicated control tasks, even in the face of sensorimotor delays, noise and perturbations -- think of the dexterity of a surgeon or even the simple task of crossing a road without getting run over. Matching this with autonomous robotics systems is challenging. Broadly speaking, challenges lie in the domain of robust sensing, flexible planning, appropriate representation and adaptive dynamics under various contexts. Statistical Machine Learning provides ideal tools to deal with these challenges, especially in tackling issues like partial observability, noise, redundancy resolution and scalability. I will illustrate some of our success stories and talk about the spills and thrills of working on exciting robotics platforms such as KUKA and SARCOS dexterous arms, Touch Bionics iLIMB and humanoid robots such as the HONDA ASIMO, DB and the Nao footballers. I will also reflect on the impact of this work on domains ranging from assisted living and autonomous navigation to prosthetics, robotic radiotherapy, exoskeletons and green energy. See you at the races!
This article was published on Feb 6, 2013