Answers are grounded in this paper and cite their sources.
This paper explores how Generative AI can transform Design for Safety & Reliability (DfSR) by proactively identifying and mitigating risks throughout the product development lifecycle. It examines three key approaches to leveraging generative AI in DfSR: Assistive Ai: This approach employs Large Language Models (LLMs) or Large Multimodal Models (LMMs) to interact dynamically with users. While highly flexible and agile, interpretability and control of AI-generated outputs remain areas of concern. Potential applications include providing diverse perspectives during brainstorming sessions. Scripted Ai: This approach uses predefined rules and scripts to automate repetitive tasks to improve the efficiency and consistency. However, its inflexibility limits its effectiveness in complex scenarios. We suggest using Scripted AI to automate routine tasks. Agentic Ai: This approach involves AI systems autonomously executing tasks, demonstrating a higher degree of “agency”. While offering real-time analysis and continuous learning capabilities, ethical considerations and system complexity present challenges. The paper highlights how these approaches can be applied across various stages of the DfSR workflow, from defining requirements and planning to risk assessment, design optimization, and sustainment. Each stage benefits from the unique strengths of each AI approach. We also emphasize the limitations of Generative AI tools. We recommend maintaining human oversight in critical decisions, treating AI as collaborators, and continuously monitoring AI systems for reliability and performance. As AI technology advances, its integration into engineering workflows promises safer, more reliable, and innovative products. Continued research and development in AI will be crucial for realizing the full potential of AI in shaping the future of DfSR.
No takes yet. Share an insight, caveat, or question.
Hu et al. (2025) studied this question.