Search Strategies for Self-driving Laboratories with Pending Experiments
This work addresses efficiency challenges in materials science automation, but it is incremental as it builds on existing Bayesian optimization methods for a specific domain.
The study tackled the problem of delayed feedback in self-driving laboratories (SDLs) due to asynchronous parallel experiments, comparing optimization strategies like expected improvement and random sampling using a simulator based on 177 real experiments to maximize conductivity of functional coatings, with results highlighting trade-offs between parallel operation and feedback delays.
Self-driving laboratories (SDLs) consist of multiple stations that perform material synthesis and characterisation tasks. To minimize station downtime and maximize experimental throughput, it is practical to run experiments in asynchronous parallel, in which multiple experiments are being performed at once in different stages. Asynchronous parallelization of experiments, however, introduces delayed feedback (i.e. "pending experiments"), which is known to reduce Bayesian optimiser performance. Here, we build a simulator for a multi-stage SDL and compare optimisation strategies for dealing with delayed feedback and asynchronous parallelized operation. Using data from a real SDL, we build a ground truth Bayesian optimisation simulator from 177 previously run experiments for maximizing the conductivity of functional coatings. We then compare search strategies such as expected improvement, noisy expected improvement, 4-mode exploration and random sampling. We evaluate their performance in terms of amount of delay and problem dimensionality. Our simulation results showcase the trade-off between the asynchronous parallel operation and delayed feedback.