Asphalt

Anyone who has driven on Indian roads knows the problem personally: potholes everywhere, and no systematic data about where they are. Cities can't fix what nobody has mapped.

Dedicated survey vehicles are expensive. Phones with accelerometers are already in every car, motorcycle, and auto rickshaw.

An Android SDK detects road anomalies on the device itself using accelerometer and gyroscope data, then a backend clusters reports from many users and exposes the findings through a map API.

The design choices reflect the actual streets it is meant for: vehicle-specific detection profiles for cars, motorcycles, and auto rickshaws; battery-aware operation that only activates above 15 km/h; offline-first buffering for patchy connectivity.

Privacy first: no personal data collection. The anomaly is the data, not the person who drove over it.

A working open source pipeline from phone sensor to city-level map, available for anyone to deploy.

Early days and honest about it, but it is the project closest to my belief that technology should serve the streets it comes from.

Constraints from the real world, like auto rickshaw suspensions and dying phone batteries, make a product design better, not worse.

Privacy-first is easier to build from day one than to retrofit, and it changes which features you even consider.