ChemOS: An Orchestration Software to Democratize Autonomous Discovery
This chapter provides an overview of established algorithmic strategies for experiment planning for closed-loop experimentation highlighting their strengths and limitations through key examples from academia and industry. It also details the need for a transition from automation to autonomy in materials innovation and process optimization to accelerate discovery across sectors. In this context, we review the early realization of autonomous laboratories, and their associated strategies to optimization, and lay out a roadmap for deploying and orchestrating self-driving laboratories. As a specific tool to enable autonomy in technology innovation, we detail the architecture and suite of applications composing the ChemOS software package. We complete our discussion by highlighting recent demonstrations of ChemOS in chemistry, materials science and process optimization and discuss the specific use of ChemOS to accelerate drug discovery.