ideasIdea
Self-driving laboratories that run the cancer biology hypothesis loop autonomously
Robotic labs guided by AI that design experiments on tumour models, run them, read the results and design the next ones, around the clock, with every result published openly.
Autonomous laboratories exist in chemistry and materials science, and cloud labs and robotic organoid culture are emerging in biology. Cancer biology is limited by slow, poorly reproducible manual experimentation. The proposal is a network of self-driving cancer labs: automated organoid and cell line culture, perturbation, imaging and sequencing readouts, active-learning experiment selection against defined questions (resistance mechanisms, combination synergy, dependency mapping), and automatic public deposition of raw data and protocols.
Hypothesis
Self-driving labs produce reproducible, reusable results at ten times the throughput and a fraction of the cost per experiment of conventional labs, and their findings translate at a higher rate.
Rationale
Automation removes the variability that undermines reproducibility and lets exploration scale with compute rather than with postdoc hours; open deposition keeps the results usable.
What would test it
Build one facility focused on resistance to KRAS inhibitors; compare throughput, replication rates and independent validation of findings against a conventional consortium over three years.
Maturity
preclinical evidence
Who has to act
engineering
Cost to try
Large (over $50M)
Years to first evidence
6
Bottlenecks it attacks
- Preclinical models that do not predict people · Nine in ten cancer drugs that work in mice fail in humans. Our models are the reason.
- Preclinical results do not reproduce · Fewer than half of landmark cancer biology findings reproduce when someone else tries.
- The valley of death between lab and product · Most academic discoveries die before anyone tests them in people because nobody funds the middle step.