<p>➀ MIT research shows AI-enabled eco-driving techniques can reduce intersection CO₂ emissions by 11-22% without impacting traffic flow, using simulations of 6,000+ U.S. intersections;</p><p>➀ A 10% adoption rate achieves 25-50% of maximum potential emission cuts, while optimizing speed limits at key intersections delivers 70% of benefits;</p><p>➀ The approach integrates deep reinforcement learning and aligns with EV adoption, offering scalable solutions for climate and public health goals.</p>