RCA 1:

Theory of Information-Energy (IE) Exchange


























Research Concentration Area (RCA) 1

In a split second, sensations in the squirrel's limbs need to be transmitted to its brain and interpreted by an internal accounting of the body's engagement with the environment, and new instructions need to be sent back to the limbs to execute the correct movements. Long before this can happen, the material properties of the animal's body have already begun to respond intelligently to the environment's mechanical stimulus. Over a much longer time scale, the animal's internal representation of where it is and what it needs to do relative to where it wants to be is adjusted as well. One of the MURI's central aims will be to establish a formal understanding of the information and energy exchange at these different timescales between the environment and the animal's integrated body-brain intelligence and to apply that insight to the construction of far more capable robots than exist today.

Conley's fundamental theorem for a class of hybrid systems

Kvalheim, Matthew D., Paul Gustafson, and Daniel E. Koditschek. "Conley's fundamental theorem for a class of hybrid systems." arXiv preprint arXiv:2005.03217 (2020).

The law of conservation of energy states that energy is neither created nor destroyed: the total energy of an isolated physical system remains constant over time. But for a non-isolated physical system, the total energy typically decreases until the system comes to rest at some steady state (unless energy is pumped into the system). For example, friction will eventually bring a swinging pendulum to a halt (unless it receives another push). Viewed more abstractly, such a physical system can be understood as moving downhill along an energy landscape until it reaches the lowest point of the "valley" to which the initial physical state belongs.

Mathematical models of physical systems such as those governed by Newton's laws are examples of the more general dynamical systems which mathematically model processes that change over time. Amazingly, in the late 1970s the mathematician C. Conley showed that all "reasonable" dynamical systems have surprisingly much in common with the physical systems mentioned above: they all have "generalized energy landscapes", and any initial state of the system evolves by moving downhill along the landscape until reaching the bottom of the valley to which it belongs. This powerful insight, established for dynamical systems evolving continuously in time, was also shown to apply to dynamical systems evolving at discrete time instants by the mathematician J. Franks in the late 1980s.

However, many physical systems—such as robots, cyber-physical systems, and pinball machines—are best modeled by dynamical systems which are hybrid in the sense that they involve both continuous-time evolution and discrete-time "events". While important contributions have been made by many researchers, the theory of such "hybrid systems" is still in its infancy, and it was unknown whether the “generalized energy landscape” insights of Conley and Franks might extend to hybrid systems. In a recent arXiv preprint [KGK20] M. D. Kvalheim, P. Gustafson, and D. E. Koditschek have shown that the insights of Conley and Franks do indeed extend to a broad class of hybrid systems relevant, e.g., for robotics. The proof of this fact in [KGK20] works by "stretching" and "gluing" the "space of all states" of the hybrid system to produce a certain continuous-time dynamical system to which Conley’s results can be applied and subsequently pulled back to the hybrid system. A more technical description of this work is given in the abstract here.

References

Koditschek, Daniel E. “The control of natural motion in mechanical systems.” Journal of Dynamic Systems, Measurement, and Control, vol. 113, pp. 547–551, 1991.

Haynes, Galen C., Fred R. Cohen, and Daniel E. Koditschek. “Gait Transitions for Quasi-static Hexapedal Locomotion on Level Ground.” in Robotics Research, C. Pradalier, R. Siegwart, and G. Hirzinger, Eds. Springer, Berlin, Heidelberg, pp. 105–121, 2011.

De, Avik, and Daniel E. Koditschek. “Parallel composition of templates for tail-energized planar hopping.” In 2015 IEEE International Conference on Robotics and Automation (ICRA), pp. 4562–4569. IEEE, 2015.

Johnson, Aaron M., and Daniel E. Koditschek. “Toward a vocabulary of legged leaping.” in 2013 IEEE International Conference on Robotics and Automation (ICRA), pp. 2568–2575. IEEE, 2013.

Revzen, Shai, and Daniel E. Koditschek. “Why We Need More Degrees of Freedom.” In Procedia IUTAM, vol. 20, pp. 89–93, 2017.

Arslan, Omur, Dan P. Guralnik, and Daniel E. Koditschek. “Coordinated robot navigation via hierarchical clustering.” IEEE Transactions on Robotics, vol. 32, no. 2, pp. 352–371, 2016.