UTCS CPS

Introduction

Cyber-Physical System Research is headed by Prof. Aloysius K. Mok. In the past few years, we have worked towards laying the groundwork for establishing a firm theoretical foundation for real-time and cyber-physical systems and also to build design tools based on this foundation.

Projects

RT-WiFi:
RT-WiFi is a real-time high-speed communication protocol for wireless cyber-physical control applications. RT-WiFi is a Time Division Multiple Access (TDMA)-based data link layer protocol built on IEEE 802.11 physical layer to provide deterministic timing guarantee on packet delivery and adjustable high sampling rates currently up to 6kHz. It incorporates configurable components for adjusting design trade-offs including sampling rate, latency variance, reliability, and compatibility to existing Wi-Fi networks, and thus can serve as a viable communication platform for supporting a wide range of high-speed wireless control systems. For more information, see RT-WiFi project.

Cyberphysical Avatar:
This project is working with the Human Centered Robotics Lab (HCRL) led by Prof. Luis Sentis and the Neural Networks Research Group (NNRG) led by Prof. Risto Miikkulainen in the University of Texas at Austin to realize a semi-autonomous robotic system called cyberphysical avatar. In contrast to teleoperated robots, a cyberphysical avatar operates semi-autonomously by complying with higher-level supervisor commands. A cyberphysical avatar can adjust to an unstructured environment and perform physical tasks subject to critical timing constraints while under human supervision. It integrates four key technologies: body-compliant control in robotics, scene understanding in computer vision, neuroevolution in machine learning, and quality-guarantees in real-time constraints.

Automated Particle Filter Design System
Automated particle filter design system is a real-time system that facilitates the design process of particle filter estimator by allowing efficient exploration of the design space, studies the statistical efficiency of the proposed particle filter estimator via simulation, and predicts total computation time when implemented on GPUs. For more information, see APFDS project.