AI for Engineering Data and Design: Balancing Assistance with Human Oversight

While AI offers significant potential for accelerating engineering tasks, particularly in design and data management, its limitations and the need for human oversight remain crucial considerations. Entrusting critical decisions entirely to AI, especially concerning data management, is not yet feasible.

The primary concern revolves around the decision-making process of AI algorithms. As Jim Schultz, product marketing manager at Synopsys, points out, “AI is making choices for you,” such as in design space optimization (DSO). While DSO can explore solutions a human might not consider, it also risks wasting computational resources if it fails to find a better outcome. This highlights the trade-off between AI’s exploratory capabilities and the potential for inefficiency.

Furthermore, the opacity of machine learning (ML) algorithms raises questions about trust. Since the decision-making process about what data to keep and discard can be unclear, many companies prefer to retain human control over these choices. Schultz notes that ultimately, data retention policies, which dictate when data should be purged, are determined by individual companies. AI can analyze data usage and make recommendations, but the final decision rests with human engineers.

Another challenge lies in the implementation of AI across different teams and data management systems. Simon Rance, general manager at Keysight, observes that data siloing can occur when different teams use disparate systems, hindering access to successive data. This necessitates a more holistic approach to engineering lifecycle management, where data can be shared securely across teams.

Startups face a unique disadvantage in leveraging AI for data management. Many AI tools require large amounts of training data, which established companies with existing product lines possess. Startups, lacking this historical data, may need to expend more resources to maintain larger amounts of data, including “bad” data, to effectively train their AI models, as Suhail Saif, principal product manager at Ansys, explains.

Additionally, over-reliance on AI is something to be on the lookout for. Despite AI’s ability to automate tasks and provide insights, it should be viewed as a supportive tool rather than a replacement for human engineers. Data engineers are needed to ensure the quality of data and manage it appropriately, as design engineers may not prioritize these tasks.

While AI is undoubtedly a valuable resource in engineering, its limitations in data management necessitate human guidance and oversight. The technology’s ability to make critical decisions about data retention, particularly in the context of complex engineering workflows, is still evolving. As AI’s capabilities advance, maintaining a balanced approach that leverages its strengths while preserving human control remains essential for ensuring efficient and reliable engineering outcomes.