AI-enhanced Systems Engineering for the Development of complex Medical Devices
summary
Claims that generative AI can cut manual documentation and speed certification by ingesting regs and producing requirements, but only if the model is decomposed into clear subtasks and backed by explicit knowledge. Prototype uses ChatGPT/Gemini with retrieval-augmented systems and structured, machine‑readable I/O, plus benchmark datasets labeled by experts. Team builds an applicability checker for MDR, runs each test multiple times, finds correct sometimes for a German medical device but wrong consistently for a US fitness app, and adds automated evaluation to isolate and fix errors.
AI-written summary · about
chapters (7)
- 0:00AI in medical device engineering
- 1:05Medical systems engineering challenges
- 3:33Generative AI capabilities
- 5:42Generative AI limitations
- 10:14Building robust AI solutions
- 13:26AI-assisted regulation checks
- 16:25Trustworthy AI in medical engineering
related talks
45mGenerative AI in medicine: Improving patient & provider experiences
15mWhat’s Next in Medical AI?
17mAI Innovation in Health
23mAI in the Life Science Sector
21mGen AI in Health: What's New in 2026?
28mThe Future of Healthcare – AI’s Promise and the Challenges Ahead
44mAI Engineering Anti-patterns
17mGenerative AI: Keeping it real