Control and Autonomy in the Age of Cyber-Physical Systems

We realize that these are exciting times for cyber-physical systems, and also that there
are several challenges that lie ahead. What are the biggest challenges in integrating computer
algorithms with mathematical models of physical (dynamical) systems? In particular, what
are the challenges in making these systems autonomous and self-reliant? Can the physics of
a system, as observed through available sensor feedback, be learnt \well enough" to provide
a good mathematical model to describe the physics? Assuming that the underlying physics
can be learnt from the observed data, can this physics be learnt in real-time or fast enough
for feedback control? When the model of the physics is already suciently known, it can
be readily incorporated in computer algorithms to learn the part that is not known (i.e.,
the uncertainties). In this talk, I argue that adaptation and learning are very important
challenges to be overcome for cyber-physical systems, where the cyber part needs to adapt
and learn from what it observes of the physical world. But also important is acceptance
of a certain amount of risk that comes with the understanding that the best mathematical
or computational model is not perfect and may not be accurate enough in all situations.
Inspiration and observations from nature, in particular from biological systems, may provide
valuable insights into how to cope with uncertainties while still achieving a desired degree
of autonomy and self-reliance in cyber-physical systems designed to carry out specific tasks.
Amit Sanyal obtained the B.Tech. degree in Aerospace Engineering from the Indian Institute of
Technology, Kanpur, in 1999. He obtained the Distinguished Graduate Student Masters Research
Award for his MS thesis in Aerospace Engineering from Texas A& M University in 2001. He
received the Ph.D. in Aerospace Engineering and his MS in Mathematics from the University of
Michigan in 2004 and 2005 respectively, where he was recipient of an Engineering Academic Scholar
Certicate. Between 2004 and 2006, he was a Post-doctoral Research Associate in the Mechanical
and Aerospace Engineering department at Arizona State University. From 2007 to spring 2010, he
was Assistant Professor in Mechanical Engineering at the University of Hawaii. Between fall 2010
and spring 2015, he was a faculty in Mechanical and Aerospace Engineering at New Mexico State
University. He is now Associate Professor in Mechanical and Aerospace Engineering at Syracuse
University. He develops and uses techniques from geometric mechanics, nonlinear and geometric
control, and variational integration of Lagrangian/Hamiltonian systems, to dynamics modeling,
guidance, navigation, and control of unmanned and autonomous systems. He is a senior member
of AIAA and IEEE, and a member of ASME and SIAM.