Tiny particles released as vehicle tires wear down may do more than pollute urban air, water, and soil. A new perspective article suggests that tire microplastics could create environmental hotspots that help antibiotic resistance genes persist, move between bacteria, and spread through cities.

Artificial intelligence could help transform how farms manage nitrogen by connecting crop production, livestock feeding, manure recycling, and environmental monitoring within a single decision-making system, according to a new perspective published in Nitrogen Cycling.

Automotive manufacturing produces large quantities of paint sludge containing volatile organic compounds, or VOCs, that can escape into the air during storage and disposal. A new study has developed a physics-based framework for predicting how temperature influences these emissions, offering practical guidance for safer hazardous-waste management.

- Professor Junmo Kim's team in the School of Electrical Engineering has developed a new AI safety verification framework called Stable-GFlowNet

- The team uncovered roughly seven times more attack types than existing methods while maintaining a 92% attack success rate

- The framework is expected to be applied not only to AI safety verification but also to a range of generative AI tasks, including drug candidate generation

Scientists have found a genetic switch that could shave years off avocado breeding by allowing researchers to identify one of a young tree’s most important traits long before it flowers.

Oak Ridge National Laboratory and Rutherford Energy Ventures recently kicked off FULCRA, an innovative consortium that combines national lab capabilities with private-sector investment to develop shared fusion testbeds, de-risk critical technology, and speed commercialization. As the first initiative of its kind, FULCRA has the potential to reshape how fusion test infrastructure is designed, funded, and operated, creating a model that can be replicated across the public research ecosystem.

The VLASH technique, developed by MIT researchers, helps robots think ahead while moving, eliminating lags that occur between different chunks of action. This smooths and streamlines robot motion, accelerating performance on tasks like pick-and-place, sorting, and stacking.