Past Event: Oden Institute Seminar

Designing intelligent machines via reactive systems

Suguman Bansal , Ph.D. candidate, Department of Rice University

Abstract

Intelligent machines, such as IoT devices, are fundamentally reactive systems that interact with the outer physical environment to achieve a goal. Asynchronous interactions are ubiquitous in reactive systems and complicate the design and programming of reactive systems. Automatic construction of asynchronous reactive programs from specifications ("synthesis") could ease the difficulty, but known methods are complex, and intractable in practice. In this talk, I will formulate a direct, exponentially more compact automaton construction for the reduction of asynchronous to synchronous synthesis. Experiments with a prototype implementation of the new method demonstrate practical feasibility. Furthermore, it is shown that for several useful classes of temporal properties, automaton-based methods can be avoided altogether and replaced with simpler Boolean constraint solving. Bio Suguman Bansal is a final-year Ph.D. candidate at Rice University advised by Prof. Moshe Y. Vardi. Her research focuses on applying formal methods to building provably verifiable AI systems. She looks at the formal reasoning of quantitative measures of systems and uses them to design tools and techniques for enhancing automated verification and synthesis of AI systems. She is the recipient of a 2019 Future Faculty Fellowship, 2018 Rising Stars Award, 2016 Andrew Ladd Fellowship. She was a visiting fellow at the Simons Institute for Theoretical Computer Science at the University of California, Berkeley in 2018. She received her M.S. in Computer Science from Rice in 2016 and B.S. in Mathematics and Computer Science from Chennai Mathematical Institute in 2014.