Charles Hartsell, Nagabhushan Mahadevan, Shreyas Ramakrishna, Abhishek Dubey, Theodore Bapty, Taylor T. Johnson, Xenofon D. Koutsoukos, Janos Sztipanovits, Gabor Karsai
Proceedings of the 30th International Workshop on Rapid System Prototyping, RSP 2019, New York, NY, USA, October 17-18, 2019 2019
CPS development with learning components faces significant challenges in integrating machine learning with formal safety assurance. This case study demonstrates how structured model-based development processes and the ALC toolchain enable developers to maintain safety guarantees while leveraging the benefits of machine learning for adaptive control. The work bridges the gap between traditional CPS design and modern learning-based approaches.
This case study paper presents the development of an autonomous Unmanned Underwater Vehicle (UUV) using Learning-Enabled Components (LECs) within the ALC (Assurance-based Learning-enabled CPS) toolchain. The work integrates system modeling, machine learning training, and verification to create a complete CPS with LEC for autonomous navigation and obstacle avoidance.
The paper demonstrates successful development of an AUV controller combining reinforcement learning and conventional control algorithms. The system achieves effective navigation through simulated underwater environments with path planning based on CNN-processed camera images. Multiple implementation alternatives are compared and evaluated using the toolchain's training and verification capabilities.
@inproceedings{Hartsell2019b,
author = {Hartsell, Charles and Mahadevan, Nagabhushan and Ramakrishna, Shreyas and Dubey, Abhishek and Bapty, Theodore and Johnson, Taylor T. and Koutsoukos, Xenofon D. and Sztipanovits, Janos and Karsai, Gabor},
booktitle = {Proceedings of the 30th International Workshop on Rapid System Prototyping, {RSP} 2019, New York, NY, USA, October 17-18, 2019},
title = {CPS} Design with Learning-Enabled Components: {A} Case Study},
year = {2019},
pages = {57--63},
abstract = {Cyber-Physical Systems (CPS) are used in many applications where they must perform complex tasks with a high degree of autonomy in uncertain environments. Traditional design flows based on domain knowledge and analytical models are often impractical for tasks such as perception, planning in uncertain environments, control with ill-defined objectives, etc. Machine learning based techniques have demonstrated good performance for such difficult tasks, leading to the introduction of Learning-Enabled Components (LEC) in CPS. Model based design techniques have been successful in the development of traditional CPS, and toolchains which apply these techniques to CPS with LECs are being actively developed. As LECs are critically dependent on training and data, one of the key challenges is to build design automation for them. In this paper, we examine the development of an autonomous Unmanned Underwater Vehicle (UUV) using the Assurance-based Learning-enabled Cyber-physical systems (ALC) Toolchain. Each stage of the development cycle is described including architectural modeling, data collection, LEC training, LEC evaluation and verification, and system-level assurance.},
bibsource = {dblp computer science bibliography, https://dblp.org},
biburl = {https://dblp.org/rec/bib/conf/rsp/HartsellMRDBJKS19},
category = {selectiveconference},
contribution = {colab},
doi = {10.1145/3339985.3358491},
file = {:Hartsell2019b-CPS_Design_with_Learning-Enabled_Components_A_Case_Study.pdf:PDF},
keywords = {autonomous underwater vehicles, learning-enabled components, reinforcement learning, model-based design, verification},
project = {cps-autonomy},
tag = {ai4cps},
timestamp = {Thu, 28 Nov 2019 12:43:50 +0100},
url = {https://doi.org/10.1145/3339985.3358491}
}