Finding Safety Violations of AI-Enabled Control Systems through the Lens of Synthesized Proxy Programs
Abstract
Given the increasing adoption of modern AI-enabled control systems, ensuring their safety and reliability has become a critical task in software testing. One prevalent approach to testing control systems is falsification, which aims to find an input signal that causes the control system to violate a formal safety specification using optimization algorithms. However, applying falsification to AI-enabled control systems poses two significant challenges: (1)~it requires the system to execute numerous candidate test inputs, which can be time-consuming, particularly for systems with AI models that have many parameters, and (2)~multiple safety requirements are typically defined as a conjunctive specification, which is difficult for existing falsification approaches to comprehensively cover. This paper introduces Synthify, a falsification framework tailored for AI-enabled control systems. Our approach performs falsification in a two-phase process. At the start, Synthify synthesizes a program that implements one or a few linear controllers to serve as a proxy for the AI controller. This proxy program mimics the AI controller's functionality but is computationally more efficient. Then, Synthify employs the $\epsilon$-greedy strategy to sample a promising sub-specification from the conjunctive safety specification. It then uses a Simulated Annealing-based falsification algorithm to find violations of the sampled sub-specification for the control system. To evaluate Synthify, we compare it to PSY-TaLiRo, a state-of-the-art and industrial-strength falsification tool, on 8 publicly available control systems. On average, Synthify achieves a 83.5% higher success rate in falsification compared to PSY-TaLiRo with the same budget of falsification trials. The safety violations found by Synthify are also more diverse than those found by PSY-TaLiRo, covering 137.7% more sub-specifications.
- Publication:
-
arXiv e-prints
- Pub Date:
- October 2024
- DOI:
- arXiv:
- arXiv:2410.04986
- Bibcode:
- 2024arXiv241004986S
- Keywords:
-
- Computer Science - Software Engineering
- E-Print:
- Under Review by ACM Transactions on Software Engineering and Methodology (TOSEM)