Margi.AI

Pothole detection built for proactive road maintenance — catching failures while they're still cheap to fix.

The problem

Roads deteriorate quietly. A crack becomes a pothole, a pothole becomes structural failure — and by the time it's reported, the fix is far more expensive than it needed to be. Digital maps don't catch this; they're built for navigation, not condition monitoring, so the data is always stale.

This isn't a small inefficiency. In the US alone, road maintenance backlogs have reached roughly $105 billion, and pothole damage cost drivers over $26 billion in a single year. Agencies are structurally set up to react to failure rather than prevent it.

The approach

Margi.AI is built to flip that model — detecting pavement deterioration early and feeding real-time condition data back to the people responsible for maintaining it, so repair happens on the cheap end of the cost curve instead of the expensive one.

Why the timing matters

Pavement economics are brutal but well understood: preservation treatments applied early can cost a fraction of what full reconstruction costs once a road has failed. The Federal Highway Administration's own guidance is explicit about this — timely preservation postpones costly rehabilitation, and the earlier the intervention, the larger the saving. The bottleneck was never the economics. It's knowing where and when to intervene, at scale, before someone files a complaint.

Where it sits

Margi.AI is the first product built under Arcanum's real-time infrastructure data thesis — proof that construction data can move at the speed the physical world actually changes, not the speed of a manual inspection cycle.