Assuring Safety-Critical Machine Learning-Enabled Systems: Challenges and Promise
نویسندگان
چکیده
We outline how assurance processes work for conventional systems and identify the primary difficulty in applying them to machine learning-enabled systems. then a path forward, identifying where considerable research remains.
منابع مشابه
Assuring Separation of Safety and Non-safety Related Systems
Safety standards call for the separation of safety and nonsafety related systems. Although good guidance is provided in these standards on how to achieve the required hazard analysis, safety integrity assignment and validation to prove a safe system, there is little available on establishing safety boundaries around the critical components and the proof of isolation from non-safety functions. D...
متن کاملDesperately Seeking Theory: Gamification, Design, Data, Machine Learning, and the Promise of Double Loop Learning Systems
Sebastian Deterding is a designer and researcher working on playful and motivational design for human flourishing. He is a senior research fellow at the Digital Creativity Labs at the University of York, and a founder and principal designer of the design agency coding conduct. He has created engaging experiences touching millions of users for clients including the BBC, BMW, Deutsche Telekom, an...
متن کاملMachine Learning for Interactive Systems: Challenges and Future Trends
Machine learning has been introduced more than 40 years ago in interactive systems through speech recognition or computer vision. Since that, machine learning gained in interest in the scientific community involved in humanmachine interaction and raised in the abstraction scale. It moved from fundamental signal processing to language understanding and generation, emotion and mood recognition an...
متن کاملSemantically-enabled Model-based Systems Engineering of Safety-critical Network of Systems
This paper describes a novel approach to the development and integration of semantics to the model-based systems engineering and operation of safety-critical network of systems. Engineering models work directly with formal domain and meta-domain (especially time and space) knowledge that are determinate, provable (ambiguity free) and executable. Knowledge is encoded as semantic blocks, which ar...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE Computer
سال: 2023
ISSN: ['1558-0814', '0018-9162']
DOI: https://doi.org/10.1109/mc.2023.3266860