CPS-PI Meeting 2018

Novel techniques in micro-fabrication are integrated with geometric-based motion planning techniques to dramatically change the fundamental capabilities in applications as diverse as noninvasive surgical procedures and nano-scale manufacturing. This integration is achieved through the development of a closed-loop process that informs mechanical design via geometric modeling, constructs prototype systems across different physical scales, and conducts experimental trials to validate results.

The increasing set of functionalities, network interoperability, and system design complexity have introduced security vulnerabilities in cyber-physical systems (CPS). As recently demonstrated, a remote attacker can disrupt the operation of a car to either disable the vehicle or hijack it. High-profile security incidents in other CPS domains include a large-scale attack on Ukraine's power-grid and the StuxNet attack on an industrial system, while the RQ-170 Sentinel drone capture has shown that even safety-critical military CPS can be compromised.

Abstract: Overall Objective.The goal of this project is to facilitate the timely retrieval of dynamic situational awareness information from field deployed nodes by an operational center in disaster recovery or search and rescue missions, which are typically characterized by resource-constrained uncertain environments. Efficient situational awareness information retrieval under severe resource limitations is critical in applications like disaster response.

o reduce reliance on other countries for minerals (e.g., coal, rare-earth metals), the USA has seen an invigoration of mining activity in recent years. Unfortunately, miners often have to work in dangerous environments where there is risk of mine explosions, fires, poisonous gases, and flooding in tunnels. Mine accidents have killed over 500 US and 40,000 mine workers worldwide in the past decade.

The goal of this proposed research project is to achieve true onboard autonomy in real time for small UAVs in the absence of remote control and external navigation aids. Three major areas have been explored. In the area of UAV flight control, an automatic trajectory generation framework is developed. It consists of waypoint planning at upper level and LQR based trajectory generation in the lower level. The Deep reinforcement learning based framework reduces the control trust by more than 15% with much less computing complexity compared to state-of-the-art approaches.