Robotic automation of Caenorhabditis elegans lifespan based on active vision

Abstract The lifespan assay in Caenorhabditis elegans is a highly repetitive and labor-intensive procedure that can lead to counting errors and delays in obtaining results. Technicians and researchers should devote their time to more intellectually demanding tasks rather than simply observing the movement of nematodes. Full automation remains a major challenge across many fields, and it is particularly beneficial in preclinical research such as lifespan assays. Developing new techniques and evaluating their efficiency is essential in order to enable the large-scale adoption of these new methods. Previous work has automated plate inspection using sequences of images of Petri dishes captured with cameras. However, integrating these computational systems into robotic platforms greatly increases the capacity for high-throughput analysis of the plates and raises new questions, such as whether these systems could affect experimental results. In this work, we present miniTower, a robotic platform for automated lifespan assays, and show that it does not alter lifespan results and, in addition, enables high-throughput automated analysis, including during holidays, without human supervision. The approach was evaluated using the low-motility strain unc-1 , in which movement detection is more challenging, making it a suitable model for validating lifespan assays.