Phone-based road condition scores for highway operators between NHAI ratings
TL;DR Ridegrade turns phones mounted in a highway operator's own patrol vehicles into a weekly road-condition check. It estimates roughness and flags likely potholes lane by lane, so crews can fix them before the official survey shows up. Highway concessionaires pay a subscription per kilometre because a poor NHAI rating can bar them from new work.
Bad road surfaces kill people. Pothole-linked crashes took 9,438 lives in India between 2020 and 2024, rising from 1,555 deaths in 2020 to 2,385 in 2024, according to road ministry data given to Parliament[1].
For highway operators the pressure is now in the contract too. Since December 2024, NHAI rates concessionaires every six months. Most of the score comes from a pavement condition index measured by survey vehicles, and anyone scoring under 70 is labelled a non-performer and can't win new national highway projects[2][3]. And between those surveys? Operators mostly rely on patrol staff spotting defects by eye.
The seed pictured the public crowdsourcing sensor data from their phones, and I've reframed it. Drivers have little reason to keep an app running and a city has no clear budget to pay for it. A highway operator does, because it already sends patrol and maintenance vehicles over every kilometre daily and its rating depends on what they miss.
Ridegrade gives those vehicles a fixed phone mount and an app that runs in the background. The phone's motion sensors and GPS record how the vehicle bounces, and a model turns that into an estimated roughness score for each 100-metre stretch, adjusted for speed and vehicle type. Sharp jolts get flagged as likely potholes, with a photo taken automatically. The operator's maintenance head sees a map that changes week to week and a list of stretches likely to drag the next rating down.
It is an early-warning tool, not a replacement for the certified survey. Research shows phone estimates are thrown off by speed, vehicle and mounting[4], so each corridor is calibrated against the latest official survey.
The concessionaire or its operations and maintenance contractor pays a subscription per lane-kilometre, probably with a setup fee for calibration. Both prices are assumptions to test. The users are patrol drivers and maintenance engineers. The alternative today is extra private surveys with laser vehicles, which cost far more per run and happen rarely.
NHAI's rating system from December 2024 put a direct penalty on poor surfaces - below 70 means no new national highway projects[2][3]. Ratings repeat every six months, so operators need to know where they stand in between. Phone-based roughness estimation has also got better, and research published in June 2024 shows data-driven methods cutting the vehicle and speed errors[4].
There is no public figure for spending on pavement monitoring by Indian highway operators. The most useful signal is how the rules are built. Every NHAI concessionaire is now rated twice a year on pavement condition, with a bar on new work below 70[2][3]. Pothole deaths[1] show why the rule exists, but they don't size this market.
One foreign startup sells a phone app that estimates road roughness and flags potholes for road agencies[5]. Network survey vehicle firms in India run the periodic laser surveys that feed NHAI's ratings. Academic teams, including one at Monash, have published smartphone methods[4]. I found no Indian product that sells continuous, between-survey condition tracking to concessionaires and ties it to their NHAI score.
| Difficulty | Moderate (38/100) |
| Starting capital | ₹45–90 L |
| Time to launch | 5–8 months |
The software is manageable. The hard part is accuracy you can defend, because phone estimates have to track the official survey closely enough that a maintenance head trusts them. You also need the official survey results for calibration, and those belong to the operator and NHAI.
No special licence here. Selling software to private highway operators needs only the baseline.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't let AI scores replace the certified survey. Phone estimates shift with speed, vehicle and mounting, so treat them as an early warning only.
Want to work out the AI layer for Ridegrade? Talk to Yash →
I'd back a founder with pavement engineering or highway operations experience, paired with a data scientist who doesn't flinch at noisy sensor data. If you want a consumer app then this is the wrong idea. First prove that phone scores on one corridor match the latest official survey well enough that the maintenance head changes a work plan.
Plan for roughly ₹45–90 L, mostly on the app, the model and calibration work with the first operators.
No sector licence. You need company registration, GST once over the threshold and DPDP compliance for drivers' location data.
Yes, wherever road operators are paid or rated on pavement condition. Check the local rating rules and your own data protection law.
The free Ridegrade validation kit has the line-by-line capital breakdown, the effort scoring, the data and consent checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for highway operators, and 5 assumptions to test. Download the kit →
Thinking about building Ridegrade or something close to it? The make-or-break bit is matching phone scores to the official survey closely enough that operators act on them, and I'm happy to help you plan that test. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.