Practical considerations for the high-level automation of a biosciences research laboratory

Abstract

Demands of research laboratories vary drastically compared to streamlined and standardised environments. This renders automating complex workflows difficult for researchers. Research environments use new and different protocols often, and for a limited number of experiments; this generates questions around the time and resource investment for automation. However, reluctance to adopt automation does not override its advantages. Automation assists the expansion of research capabilities, allows additional time to focus on planning and analysis, and even prevents shutdowns due to unforeseen circumstances. This work explores the implementation of automation circumventing coding requirements for the end-user in a bioscience research laboratory with a focus on resource efficiency. Automated operation was compared against its human counterpart to demonstrate the benefits of the no-code automation: operation time (up to 66 % reduction), errors in execution (no errors), ease of protocol modification (against code-dependent robotic platforms), consumable usage (repeated virtual simulations to minimise waste), and quality of results (minimal coefficient of variation and high correlation). This is the first work that demonstrates the implications of adopting no-code (high-level) automation into perpetually evolving research environments. This work demonstrates that widening researcher participation in automation increases experimental efficiency, makes challenging experimental designs attainable, and builds a robust laboratory environment.