(November Series #1) Search strategies for asynchronous parallel self-driving laboratories with pending points
Recording is available on Youtube now:
Hao Wen will discuss our recent NeurIPS workshop paper on "Search strategies for asynchronous parallel self-driving laboratories with pending points".
Time: Nov 8, 2023 02:00 PM London
Location: University of Surrey, Guildford, School of Chemistry Engineer
Abstract: 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 points”), 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 Rupnow et al., 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, 4-mode exploration as proposed by the original authors and asynchronous batching. We evaluate their performance in terms of number of stages, and short, medium and long-term optimisation performance. Our simulation results showcase the trade-off between the asynchronous parallel operation and delayed feedback.
The OptApp link to the OptStore will be available here early 2024!
Get in touch
We will happily guide you through the emerging space of data-driven material discovery. We look forward to learn from your experience and problem space.
