Draft:Operational assurance

  • Comment: This appears to be a simple definition of a term used by some organisations like NASA, there's no indication it as a topic has actually been covered. Devonian Wombat (talk) 07:44, 12 May 2026 (UTC)

Operational assurance is an approach to maintaining confidence that a deployed system continues to operate within intended bounds during real-world use. The term is used in connection with autonomous, safety-critical, and software-defined systems, where design-time validation may be insufficient because systems can change after deployment through updates, environmental variation, or adaptive behavior.[1][2]

NASA's technology taxonomy uses the phrase operational assurance of autonomous systems for confirming, before or during operations, that an autonomous system is operating safely and efficiently and is not adversely affecting other systems.[3] Related literature connects operational assurance with runtime monitoring, post-deployment verification, and dynamic assurance cases that extend assurance activity beyond initial certification or release.[4][5]

Overview

Operational assurance addresses the problem that systems judged acceptable at release may not remain acceptable in operation. This is especially relevant for autonomous and AI-enabled systems, where traditional assurance methods may struggle with complexity, uncertainty, and post-deployment change.[6][7]

In practice, the concept overlaps with continuous monitoring, anomaly detection, maintenance of operational constraints, and collection of evidence about deployed behavior.[8][9]

Relationship to runtime assurance

Operational assurance overlaps with runtime assurance, a field that combines design-time analysis with runtime mechanisms intended to preserve required safety or correctness properties during operation. In robotics research, runtime assurance has been described as using runtime monitors and control switching to keep systems within specified safety properties when high-performance controllers cannot be fully trusted on their own.[10]

Post-deployment monitoring

Public guidance on AI assurance has increasingly emphasized post-deployment monitoring. NIST has argued for a "trust but verify continuously" approach and its AI RMF Playbook discusses real-time monitoring, anomaly detection, incident response, and continuous feedback during deployment and operation.[11][12][13]

A 2026 NIST report also noted that terminology and best practices for monitoring deployed AI systems remain immature and fragmented, which reflects the still-emerging nature of the field.[14]

See also

References

  1. ^ National Aeronautics and Space Administration (2020). 2020 NASA Technology Taxonomy (PDF) (Report). NASA. Retrieved 2026-03-12.
  2. ^ Laplante, Phillip; Kuhn, D. Richard (2022). "AI Assurance for the Public -- Trust but Verify, Continuously". 2022 IEEE 29th Annual Software Technology Conference (STC). National Institute of Standards and Technology. doi:10.1109/STC55697.2022.00032. Retrieved 2026-03-12.
  3. ^ National Aeronautics and Space Administration (2020). 2020 NASA Technology Taxonomy (PDF) (Report). NASA. Retrieved 2026-03-12.
  4. ^ Neogi, Natasha; Young, Steven; Dill, Evan (2022). "Establishing the Assurance Efficacy of Automated Risk Mitigation Strategies". AIAA Aviation 2022 Forum. NASA Technical Reports Server. Retrieved 2026-03-12.
  5. ^ Denney, Ewen; Menzies, Jonathan; Pai, Ganesh (2023-05-11). Dynamic Assurance Cases: Closing the Loop Between Design and Operational Assurance. Software Certification Consortium Meeting 21. NASA Technical Reports Server. Retrieved 2026-03-12.
  6. ^ Neogi, Natasha; Young, Steven; Dill, Evan (2022). "Establishing the Assurance Efficacy of Automated Risk Mitigation Strategies". AIAA Aviation 2022 Forum. NASA Technical Reports Server. Retrieved 2026-03-12.
  7. ^ Fisher, Michael; Mascardi, Viviana; Rozier, Kristin Yvonne; Schiavo, Luca; Winikoff, Michael; Yorke-Smith, Neil (2021). "Towards a Framework for Certification of Reliable Autonomous Systems". Autonomous Agents and Multi-Agent Systems. 35 (8). doi:10.1007/s10458-020-09487-2. Retrieved 2026-03-12.
  8. ^ "Measure". NIST AI RMF Playbook. National Institute of Standards and Technology. Retrieved 2026-03-12.
  9. ^ Rao, Anita; Keller, Andrew; Kalra, Neha; Steed, Ryan; Kwegyir-Aggrey, Kweku; Klyman, Kevin; Staheli, Diane; Bergman, Amanda (2026). Challenges to the Monitoring of Deployed AI Systems: Center for AI Standards and Innovation (Report). NIST Trustworthy and Responsible AI. National Institute of Standards and Technology. doi:10.6028/NIST.AI.800-4. Retrieved 2026-03-12.
  10. ^ Desai, Ankush; Ghosh, Shromona; Seshia, Sanjit A.; Shankar, Natarajan; Tiwari, Ashish (2018). Soter: Programming Safe Robotics System using Runtime Assurance (Report). EECS Department, University of California, Berkeley. Retrieved 2026-03-12.
  11. ^ Laplante, Phillip; Kuhn, D. Richard (2022). "AI Assurance for the Public -- Trust but Verify, Continuously". 2022 IEEE 29th Annual Software Technology Conference (STC). National Institute of Standards and Technology. doi:10.1109/STC55697.2022.00032. Retrieved 2026-03-12.
  12. ^ "Measure". NIST AI RMF Playbook. National Institute of Standards and Technology. Retrieved 2026-03-12.
  13. ^ "Manage". NIST AI RMF Playbook. National Institute of Standards and Technology. Retrieved 2026-03-12.
  14. ^ Rao, Anita; Keller, Andrew; Kalra, Neha; Steed, Ryan; Kwegyir-Aggrey, Kweku; Klyman, Kevin; Staheli, Diane; Bergman, Amanda (2026). Challenges to the Monitoring of Deployed AI Systems: Center for AI Standards and Innovation (Report). NIST Trustworthy and Responsible AI. National Institute of Standards and Technology. doi:10.6028/NIST.AI.800-4. Retrieved 2026-03-12.

Content Disclaimer

Informasi ini disarikan dari Wikipedia dan disajikan kembali untuk tujuan edukasi. Konten tersedia di bawah lisensi CC BY-SA 3.0. Kami tidak bertanggung jawab atas ketidakakuratan data yang bersumber dari kontribusi publik tersebut.

  1. The information displayed on this website is sourced in part or in whole from Wikipedia and has been adapted for the purpose of restating it. We strive to provide accurate and relevant information, however:
  2. There is no guarantee of absolute accuracy. Wikipedia is an open, collaborative project that can be edited by anyone, so information is subject to change.
  3. It is not intended to constitute professional advice. The content displayed is for informational and educational purposes only. For important decisions (e.g., medical, legal, or financial), please consult a professional.
  4. Content copyright. Wikipedia is licensed under the Creative Commons Attribution-ShareAlike License (CC BY-SA). This means that content may be reused with appropriate attribution and shared under a similar license.
  5. Responsible use. Any risk arising from the use of information from this website is entirely the responsibility of the user.