Researchers at the Massachusetts Institute of Technology have developed a new artificial intelligence approach that works like a “ChatGPT for spreadsheets,” enabling engineers to solve complex design and optimization problems more quickly, tells MIT News. The system helps analyze massive design spaces by identifying the most important variables and focusing computational effort on the factors that most strongly influence performance.
Engineering systems such as power grids, transportation networks, and vehicle designs often involve hundreds of design parameters. Each parameter can influence performance metrics such as safety, efficiency, or cost. Engineers typically explore these parameters using optimization algorithms, but evaluating every possible combination can become computationally overwhelming. Even with modern computing resources, searching such large design spaces may require enormous time and energy.
The new MIT approach uses machine learning to guide this process. Instead of treating every design variable as equally important, the algorithm learns to identify which parameters have the strongest impact on a design goal. It then concentrates its analysis on those influential factors while reducing attention to less significant variables. This strategy allows engineers to narrow the search space and reach promising solutions more efficiently.
The method works in a way similar to a spreadsheet tool augmented with conversational AI. Engineers can explore different scenarios, adjust variables, and allow the system to highlight the combinations that are most likely to produce optimal results. By focusing on the key drivers of performance, the algorithm dramatically accelerates the exploration of complex engineering tradeoffs.
The researchers demonstrated the system on several challenging design problems, including power-grid management and vehicle design optimization. In these scenarios, the tool quickly identified the parameters most responsible for improvements in safety or efficiency, allowing engineers to reach high-quality solutions faster than conventional methods.
Beyond these examples, the researchers believe the approach could transform engineering design workflows that rely heavily on spreadsheet-based models. By combining optimization algorithms with AI-driven reasoning, the system offers a new way to manage complex, high-dimensional engineering problems while reducing computational cost and development time.