ARCH-COMP25 Category Report: Artificial Intelligence and Neural Network Control Systems (AINNCS) for Continuous and Hybrid Systems Plants

51 pages•Published: December 22, 2025

Diego Manzanas Lopez, Matthias Althoff, Luis Benet, Samuel Coogan, Marcelo Forets, Akash Harapanahalli, Taylor T. Johnson, Tobias Ladner, Christian Schilling, Huan Zhang and Xiangru Zhong

Abstract

This report presents the results of a friendly competition for formal verification of continuous and hybrid systems with artificial intelligence (AI) components. Specifically, machine learning (ML) components in cyber-physical systems (CPS), such as feedforward neural networks used as feedback controllers in closed-loop systems, are considered, which is a class of systems classically known as intelligent control systems, or in more modern and specific terms, neural network control systems (NNCS). We broadly refer to this category as AI and NNCS (AINNCS). The friendly competition took place as part of the workshop Applied Verification for Continuous and Hybrid Systems (ARCH) in 2025. In this edition of the AINNCS category at ARCH-COMP, five tools have been applied to solve 12 benchmarks, which are CORA, CROWN-Reach, immrax, JuliaReach, and NNV. For the second year in a row, we have the largest interest in the community, with four previous participants and one new participant, immrax. In
reusing the hardware infrastructure and benchmarks from last year, we can observe comparable results from previous improvements, with slight improvements in computation time by CORA and NNV in selected benchmarks. A novelty of this year is the different problem abstraction between immrax and the rest of tools, leading to result disparities in 2 benchmarks: Single Pendulum and Attitude Control.

Keyphrases: control system, neural feedback loop, neural network control system, verification

In: Goran Frehse and Matthias Althoff (editors). Proceedings of 12th Int. Workshop on Applied Verification for Continuous and Hybrid Systems, vol 108, pages 71-121.

Links:	https://easychair.org/publications/paper/Gc39
	https://doi.org/10.29007/9vg6

BibTeX entry

@inproceedings{ARCH25:ARCH_COMP25_Category_Report,
  author    = {Diego Manzanas Lopez and Matthias Althoff and Luis Benet and Samuel Coogan and Marcelo Forets and Akash Harapanahalli and Taylor T. Johnson and Tobias Ladner and Christian Schilling and Huan Zhang and Xiangru Zhong},
  title     = {ARCH-COMP25 Category Report: Artificial Intelligence and Neural Network Control Systems (AINNCS) for Continuous and Hybrid Systems Plants},
  booktitle = {Proceedings of 12th Int. Workshop on Applied Verification for Continuous and Hybrid Systems},
  editor    = {Goran Frehse and Matthias Althoff},
  series    = {EPiC Series in Computing},
  volume    = {108},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2398-7340},
  url       = {/publications/paper/Gc39},
  doi       = {10.29007/9vg6},
  pages     = {71-121},
  year      = {2025}}

Download PDF Open PDF in browser