General Motors

Solution

General Motors built an in-house AI-assisted site-selection workflow to decide where new public EV chargers should be added across the United States. The implementation starts with EV traffic-pattern data and charger network context, which GM data scientists feed into internal AI tools. Predictive analytics and geospatial algorithms then process that data to evaluate candidate sites as a mathematical optimization problem, scoring locations against criteria such as traffic flow and proximity to existing chargers. The system also produces detailed maps so teams can inspect why locations were recommended. GM does not name a storage platform or external AI vendor, but it does describe a human-in-the-loop decision process in which model recommendations are combined with expert insights before final approval. Approved opportunities are then handed off to deployment partners EVgo and Pilot Flying J to turn ranked locations into operational fast-charging sites.

Data flow

EV traffic-pattern data and charger-location context are fed into GM's AI tools, where predictive analytics, geospatial algorithms, and an optimization model score candidate sites using traffic flow and proximity criteria. The system outputs detailed maps and ranked recommendations, which GM experts review before approved opportunities are passed to EVgo and Pilot Flying J for charger deployment.