HKUST Researchers Develop Optimization Framework for Automated Materials Labs

Researchers at The Hong Kong University of Science and Technology (HKUST) have developed a computational framework designed to help engineering teams plan modularized autonomous experimentation platforms before purchasing hardware. The method determines optimal equipment configurations prior to construction, aiming to reduce setup delays and costs during laboratory assembly, according to a report by TechXplore.

Detailed in a study published in IEEE Transactions on Automation Science and Engineering, the project was co-led by HKUST mechanical and aerospace engineering professor Yang Jinglei and assistant professor Duan Molong, alongside co-first authors Ma Guoxiong and Cui Haozhe.

Rather than relying on intuitive or trial-and-error purchasing decisions, the methodology provides a mathematical basis for equipment selection. The system uses hybrid automata—a formalism for describing systems that move through discrete states over time—to map experimental workflows into procedure states, transitions, timing logic, and instrument demands. It simultaneously catalogs device specifications in a structured equipment dictionary covering operating parameters, physical footprint, equipment costs, and inter-device dependencies.

To identify feasible configurations, the team formulated hardware selection as a constrained integer nonlinear optimization problem. The formulation balances financial budgets, physical footprint limits, throughput targets, and hardware dependencies. An integer-coded differential evolution algorithm then explores the discrete decision space to find cost-effective setups that satisfy experimental requirements.

The researchers validated the method through a desktop-scale thermal insulation coating case study conducted on a 0.54-square-meter footprint with a budget of roughly $11,000. In a three-sample test batch comparing automated and manual operations, active operator time fell by 79.1%, decreasing from 1,589.7 seconds to 331.8 seconds. In addition, the coating's light transmittance demonstrated a smaller standard deviation, indicating greater process consistency.

According to Yang, the framework establishes a structured mechanism to analyze workflow timing, equipment utilization, and practical constraints before committing capital. Duan noted that shifting from experience-driven planning to quantitative design helps teams identify equipment bottlenecks in advance to build compact, high-throughput platforms.

The HKUST team plans to extend the framework to include layout-aware configuration, dynamic scheduling, and artificial intelligence for closed-loop experimental decisions, serving as an early step toward a broader software architecture they describe as a Materials Operating System (MaterOS).