Highways£2,500/mo
Winter-Grit AI
Challenge
Winter service planning often relies on static gritting routes that cannot adapt quickly to hyperlocal temperature variation, moisture, and late weather shifts. This leads to overtreatment in low-risk corridors and delayed treatment in high-risk stretches where black ice forms first.
Solution
Winter-Grit AI fuses live weather feeds, road-temperature models, and Mapbox network intelligence to generate optimal pre-ice treatment plans in near real time. Supervisors can dispatch safer, shorter, and better-timed gritting runs while evidencing treatment decisions for governance and claims defence.
Key Features
Forecast ingestion pipeline evaluates air temperature, road-surface predictions, and precipitation probability at route-segment granularity
Constraint-aware optimiser uses Mapbox graph costs to compute efficient salting paths by depot, spreader capacity, and shift windows
Adaptive replanning triggers route updates when nowcast conditions drift, with full audit trail of why each revision was made