Stormward

A utility has to commit crews days before anyone knows where a storm lands. Mutual aid is booked on a single forecast track, and both mistakes are expensive: over-mobilise and you pay full rate for crews you did not need, wave it off and you are caught with hundreds of thousands of meters dark. Stormward turns that call into a number you can defend to a COO before landfall and to a regulator afterwards.

What it answers

• Which utilities are in the path — service territories intersected with the modelled damaging-wind swath, ranked by exposure concentration rather than raw headcount.
• How many customers go dark — a calibrated fragility curve on peak sustained wind, weighted by where customers actually live rather than smeared over the territory.
• How long restoration takes — a crew-day model run against each utility’s own workforce plus capped mutual-aid mobilisation, returning an ETR and the crew shortfall.
• What it costs — restoration dollars driven by the blended damaging wind, so a reconnection event is not priced as a rebuild.
• How wrong it has been — every historical storm ships with its modelled figures against the reported outage record.

Built on real data

Track and intensity come from the National Hurricane Center’s own forecast advisories, not an extrapolation of the current motion vector. The wind field is a Holland parametric profile with forward-motion asymmetry, evaluated as the peak sustained wind each point sees across the whole passage.