University of California, Riverside researchers on Sept. 24, 2026, described a machine-learning method that screens chemical structures for odors that safely push honey bees away from pesticide-treated crops, UC Riverside News (opens in new tab) and the journal eLife (opens in new tab) reported.
Led by molecular biologist Anandasankar Ray with entomologist Boris Baer’s lab, the team trained a model on chemical structures and bee behavioral responses, then screened more than 50 million compounds and identified about 130 predicted bee-repellent candidates, the university said.
Lab avoidance tests matched model predictions; subsequent field experiments with freely foraging bees found that all seven compounds tested reliably repelled bees from honey combs without harming them, according to UC Riverside and the eLife (opens in new tab) paper titled on machine learning of honey bee olfactory behavior.
Researchers framed the work as a path toward bee-friendlier pesticide formulations and other settings where reducing bee contact matters. The Event Log treats the institutional announcement and peer-journal report as confirmed science coverage; commercial product rollout is not claimed here.