Driving · 2026-10-30
Assuring Learning-Enabled Cyber-Physical Systems with Foundation Models: A Neuro-Symbolic Agenda for Testing, Verification, and Runtime Assurance
Xi Zheng, Ivan Ruchkin, Hong Jin Kang, Tevfik Bultan, Vijay Ganesh, Taylor T. Johnson, Ziyang Li, Lei Ma, Aloysius K. Mok, Daniel Neider, Ruzica Piskac, Nalini Venkatasubramanian, Simon Thompson, Sanjoy Baruah
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
The authors argue for system-level assurance for learning-enabled cyber-physical systems that use foundation models. They propose a “Foundation Model Assurance Stack” combining (i) neuro-symbolic architectures with symbolic world models and contract specifications, (ii) conformance-based, scenario-driven testing with foundation-model-assisted specification mining and synthesis, and (iii) compositional verification plus runtime assurance using monitors and enforcement via a Simplex-style layer.
Why it matters
The paper targets a core gap: strong benchmark performance of individual foundation models does not guarantee correct behavior of the integrated CPS. The proposed stack aims to restore engineering control by creating explicit assurance boundaries between stochastic learning and formal reasoning, and by connecting testing, verification, and runtime evidence across the system lifecycle for deployable, certifiable learning-enabled CPS.
Method
- Neuro-symbolic CPS architecture using explicit symbolic world models and explicit neural–symbolic interface + assume–guarantee contract specifications to expose semantics, uncertainty, and degradation handling.
- Conformance-based testing over structured assurance artefacts: semantically grounded scenario generation and DSL-guided test synthesis, with foundation-model-assisted specification mining and adequacy assessment.
- Compositional verification and runtime assurance: contract/interface/transformation verification, runtime monitor synthesis, safety constraint synthesis, and enforcement via a Simplex-style resilience layer with a learning controller and verified safe controller/shielding or switching.
Abstract (from arXiv)
Learning-enabled cyber-physical systems (CPS), including autonomous vehicles and UAVs, increasingly incorporate foundation models, yet fragmented assurance practices provide insufficient support for system-level guarantees. We argue that restoring engineering control across stochastic neural, symbolic, and physical boundaries requires a shift from model-centric evaluation to system-level assurance. We introduce the Foundation Model Assurance Stack, connecting three directions: (i) neuro-symbolic architectures that constrain foundation models through explicit symbolic world models, reasoning layers, and interfaces; (ii) semantically grounded, conformance-based testing that combines formal specifications and symbolic world models with foundation-model-assisted specification mining, scenario generation, and test synthesis; and (iii) compositional verification and runtime assurance that target formally tractable components and interfaces through modular contracts, equivalence checking, verified synthesis, uncertainty-aware reasoning, and runtime monitoring and enforcement. Symbolic abstractions establish assurance boundaries between stochastic learning and formal reasoning, connecting testing, verification, and runtime evidence across the system lifecycle. Foundation models serve as constrained assistants whose outputs are independently checked wherever correctness can be formally specified. We identify research challenges in specification engineering, abstraction fidelity, evidence composition, and lifecycle certification, and outline a staged agenda toward trustworthy, deployable, and certifiable learning-enabled CPS.
Related papers
- An Empirical Study of Architectural Shift from Traditional to AI-Enabled Simulink Controllers
- DreamTest: World-Model Surrogates for Search-Based Testing of Deep Reinforcement Learning Agents
- CyberWorld: World Models for Sample-Efficient Autonomous Cyber Defense
- Recursive Self-Improvement in Physical AI: Mechanisms, Evidence, and Evaluation
- LOGIV: Graph-Verified Execution with Causal Local Repair for Long-Horizon Robot Tasks
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First announced by arXiv: 2026-10-30 · Indexed by Pub-AI: 2026-10-30 UTC
Assuring Learning-Enabled Cyber-Physical Systems with Foundation Models: A Neuro-Symbolic Agenda for Testing, Verification, and Runtime Assurance
First announced by arXiv: 2026-10-30 · Original record ↗
- arxiv
- 2610.39547
Explicit source versions (1)
- v1 · Date not supplied
- Preprint available
Source date: 2026-10-30 · First detected: 2026-10-30 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · RSS guid=oai:arXiv.org:2610.39547v1; arxiv:announce_type=new; item pubDate (official metadata)
Abstract · arxiv
Learning-enabled cyber-physical systems (CPS), including autonomous vehicles and UAVs, increasingly incorporate foundation models, yet fragmented assurance practices provide insufficient support for system-level guarantees. We argue that restoring engineering control across stochastic neural, symbolic, and physical boundaries requires a shift from model-centric evaluation to system-level assurance. We introduce the Foundation Model Assurance Stack, connecting three directions: (i) neuro-symbolic architectures that constrain foundation models through explicit symbolic world models, reasoning layers, and interfaces; (ii) semantically grounded, conformance-based testing that combines formal specifications and symbolic world models with foundation-model-assisted specification mining, scenario generation, and test synthesis; and (iii) compositional verification and runtime assurance that target formally tractable components and interfaces through modular contracts, equivalence checking, verified synthesis, uncertainty-aware reasoning, and runtime monitoring and enforcement. Symbolic abstractions establish assurance boundaries between stochastic learning and formal reasoning, connecting testing, verification, and runtime evidence across the system lifecycle. Foundation models serve as constrained assistants whose outputs are independently checked wherever correctness can be formally specified. We identify research challenges in specification engineering, abstraction fidelity, evidence composition, and lifecycle certification, and outline a staged agenda toward trustworthy, deployable, and certifiable learning-enabled CPS.