Virtual Labs is automating the world's most manual scientific workflow
Virtual Labs is automating the world's most manual scientific workflow. Two forces are colliding in a multi-trillion dollar industry. Regulators now demand that drug testing use fewer animals - it's written into funding criteria and submission guidance. But the same regulators still require behavioral data from animal studies, and that data is collected today by a technician with a clipboard. Virtual Labs sells the resolution to that collision: smart cameras that automate ANY legacy behavioral assay, while capturing radically more data per animal, and AI that labels that data in real time and turns it into virtual animals, so every study needs fewer real ones. Meet PolarEye(tm), the smart camera that sees what the clipboard and pencil misses: - Optical (hardware based) background removal: maze walls, reflections, animal chambers and lab clutter vanish in the optics, leaving nothing but the animal. - Real-time pose estimation: every paw, snout, and tail tracked as the behavior happens. - Facial expression analysis: objective readouts of pain, affect, and wellbeing. - Standardized behavioral phenotyping: the same quantitative phenotype in every lab and every study. - Closed-loop experiments: sync and trigger microscopy, e-phys, and other equipment in real time. Virtual Labs was founded by David Roberson, PhD, MBA, a neuroscientist and serial biotech founder who has spent more than 15 years using computer vision to study rodent behavior. David earned his PhD in neuroscience at Harvard Medical School, in the laboratory of Clifford Woolf, where his dissertation focused on analysis of rodent behavior to interpret brain function. He was named a 2014 Harvard Horizons Scholar and received Tech Titans' 2025 Technology Inventor Award. Still scoring behavior by hand? Reach out for more info on how our PolarEye technology can reliably automate scoring for over 200 standard rodent behavioral assays, including the one that only that one technician in your lab can do right.