MIT researchers create system that helps AI solve complex planning problems

MIT researchers have developed a groundbreaking framework that enables large language models (LLMs) to tackle complex planning problems by breaking them down into manageable components and leveraging specialized optimization software. Unlike previous attempts to enhance LLMs’ inherent planning capabilities, this approach guides the AI to formulate problems like humans would, then uses powerful solvers to find optimal solutions. This innovation bridges the gap between natural language interfaces and sophisticated planning algorithms, potentially transforming how businesses approach supply chain management, factory scheduling, and other complex optimization challenges.

The big picture: MIT’s new framework allows users to describe complex planning problems in plain language, which an LLM then translates into a format solvable by specialized optimization software.

• The system achieved an 85 percent success rate across nine complex planning challenges, more than doubling the 39 percent success rate of the best baseline approach.
• Unlike other approaches, this framework requires no task-specific examples or specialized prompting to train the language model.

How it works: The system breaks down planning problems through a multi-step process where the LLM formulates the problem, checks its work, and corrects errors before passing it to an optimization solver.

• Users only need to describe their planning challenge in natural language—like optimizing a coffee company’s supply chain—without needing technical expertise.
• During formulation, the LLM performs self-checks at intermediate steps and can fix errors it identifies rather than abandoning the task.

Why this matters: Traditional LLMs like ChatGPT often perform poorly when directly tasked with solving complex planning challenges that involve multiple variables and constraints.

• This approach leverages the complementary strengths of LLMs (understanding natural language) and specialized optimization software (solving complex mathematical problems).
• The versatility of the framework makes it applicable to diverse planning scenarios from warehouse robot routing to airline crew scheduling.

In plain English: Rather than trying to make AI models better at solving complex planning problems directly, MIT researchers created a system where the AI acts as an interpreter—it takes your explanation of a problem, translates it into a language that specialized planning software can understand, and then lets that software find the optimal solution.