Assured Runtime Monitoring and Planning: Toward Verification of Neural Networks for Safe Autonomous Operations
- 15 April 2020
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics & Automation Magazine
- Vol. 27 (2) , 102-116
Abstract
Autonomous systems operating in uncertain environments under the effects of disturbances and noises can reach unsafe states even while using finetuned controllers and precise sensors and actuators. To provide safety guarantees on such systems during motion planning operations, reachability analysis (RA) has been demonstrated to be a powerful tool. RA, however, suffers from computational complexity, especially when dealing with intricate systems characterized by high-order dynamics, making it hard to deploy for runtime monitoring.Keywords
Funding Information
- Air Force Research Laboratory (FA8750-18-C-0090)
This publication has 16 references indexed in Scilit:
- Methods for interpreting and understanding deep neural networksDigital Signal Processing, 2018
- Funnel libraries for real-time robust feedback motion planningThe International Journal of Robotics Research, 2017
- Robust online motion planning via contraction theory and convex optimization2017
- Exact and efficient Hamilton-Jacobi guaranteed safety analysis via system decomposition2017
- Wind Estimation in the Lower Atmosphere Using Multirotor AircraftJournal of Atmospheric and Oceanic Technology, 2017
- Reach-avoid problems with time-varying dynamics, targets and constraints2015
- Flow*: An Analyzer for Non-linear Hybrid Systems2013
- Minimum snap trajectory generation and control for quadrotors2011
- Approximate bisimulation relations for constrained linear systemsAutomatica, 2007
- Statistics of wind direction and its incrementsPhysics of Fluids, 2000