← All Publications

Conference Paper · 2026 · Accepted · forthcoming

Defining Operational Conditions for Safety-Critical AI-Based Systems from Data

Johann Maximilian Christensen, Elena Hoemann, Frank Köster, Sven Hallerbach

Advances in Neural Information Processing Systems (NeurIPS)

Workshop Acceptances

This work has also been accepted to two workshops at NeurIPS 2026 in Sydney:

The Short Version

A data-driven method reconstructs operational design domains using a deterministic kernel-based representation, with validation through Monte Carlo experiments and an aviation collision-avoidance use case.

Abstract

Artificial Intelligence (AI) has been on the rise in many domains, including numerous safety-critical applications. However, for complex systems in the real world, defining the underlying environmental conditions in which the AI-based system must operate—the Operational Design Domain (ODD)—is extremely challenging. This often results in an incomplete description of the ODD, which contrasts with the requirements of many domains for certifying AI-based systems. Traditionally, the ODD is created in the early stages of the development process, drawing on sophisticated expert knowledge and related standards. This paper presents a novel Safety-by-Design method to a posteriori define the ODD from previously collected data using a multi-dimensional kernel-based representation. This approach is validated through both Monte Carlo methods and a real-world aviation use case for a future collision-avoidance system. Moreover, by defining under what conditions two ODDs are similar, the paper shows that the data-driven ODD can produce a dataset similar to the original, hidden ODD. Deriving the novel, Safety-by-Design, deterministic kernel-based affinity representation of ODDs is fully automated via a bounded, order-independent algorithm. Utilizing the proposed ODD representation enables future certification of data-driven, safety-critical AI-based systems.

Reference

Cite This Work

BibLaTeX

@inproceedings{Christensen2026,
  author = {Christensen, Johann Maximilian and Hoemann, Elena and K{\"{o}}ster, Frank and Hallerbach, Sven},
  title = {Defining Operational Conditions for Safety-Critical AI-Based Systems from Data},
  date = {2026-12},
  booktitle = {Advances in Neural Information Processing Systems},
  volume = {40},
  pages = {1--39},
  publisher = {Curran Associates},
  location = {Red Hook, NY, USA},
  pubstate = {accepted},
  eventtitle = {NeurIPS},
  eventdate = {2026-12-06/2026-12-12},
  venue = {Sydney, NSW, Australia},
}

BibTeX

@inproceedings{Christensen2026,
  author = {Christensen, Johann Maximilian and Hoemann, Elena and K{\"{o}}ster, Frank and Hallerbach, Sven},
  title = {Defining Operational Conditions for Safety-Critical AI-Based Systems from Data},
  booktitle = {Advances in Neural Information Processing Systems},
  volume = {40},
  pages = {1--39},
  publisher = {Curran Associates},
  address = {Red Hook, NY, USA},
  year = {2026},
  month = dec,
  note = {Accepted; forthcoming},
}