Acoustic leak surveys for 24x7 water zones, paid per verified leak
TL;DR Leakfinder finds hidden leaks in pressurised water networks. Crews lay acoustic loggers across a metered zone overnight, software ranks the suspect spots and a technician pinpoints each one before the utility digs. City utilities, their contractors and large campuses pay per kilometre surveyed and per leak confirmed.
Indian cities don't know how much water they lose. In Gujarat, large cities reported non-revenue water of 19–24% in 2019, but the researchers who compiled the data called none of those figures reliable, and a preliminary audit in Kalol found 46%[1]. Only 43 urban local bodies had flow meters at their treatment plants[1], so most losses are guessed, not measured.
The national programme now wants numbers. AMRUT 2.0 sets 20% as the ideal non-revenue water level, asks cities for a roadmap that includes cutting distribution leakage and says it will judge how well each city's leak-detection system works[2].
Most utilities have no team that does this every week.
Leakfinder works one district metered area at a time. A crew places correlating acoustic loggers on valves and hydrants across the zone in the evening. Overnight, when demand is low, the loggers record the hiss of water escaping under pressure. Software compares readings between pairs of loggers and ranks the likely leak spots by pipe material and noise pattern, and next morning a technician with a ground microphone pinpoints each one. The utility gets a map, a dig list and a report it can show under AMRUT. Leakfinder gets paid when a dig confirms the leak.
I reframed the seed, which pitched always-on AI listening across a city. Leaks are silent when pipes are empty, and many Indian networks run for about two hours a day[1]. Permanent monitoring only pays off where pressure is continuous - AMRUT's 24x7 zones of 2,000 households or more[2], and private networks in industrial parks, airports and large campuses. The software helps with ranking but field crews still find the leaks.
The buyer is the urban local body or water board, or the contractor running a 24x7 supply project if its contract carries a loss-reduction target (an assumption to check in tender documents). Campus and industrial-park facility heads are a faster second buyer. I found no public price for leak surveys in India, so test a per-kilometre fee plus a bonus per confirmed leak.
AMRUT 2.0 was launched on 1 October 2021 with ₹2,99,000 crore and makes cutting non-revenue water a core reform[2]. Pey Jal Survekshan scores 485 cities, with separate 100-point indicators for measuring non-revenue water and for flow measurement[3]. Its Technology Sub-Mission funds start-up pilots up to ₹20 lakh with a state committee's approval[2].
Nobody publishes a figure for India's leak-detection services market. The narrowest anchor is demand created by policy - 485 cities ranked partly on non-revenue water[3], and a reform that ties funding to it[2]. The AMRUT 2.0 outlay of ₹2,99,000 crore is programme money for all water and sewer work, so treat it as context, not your market.
One European leak-detection maker sells correlating loggers and says a 5,500-logger rollout in Grand Lyon helped cut losses from 21% to under 15%. It sells in India through a distributor[4]. A UK startup sells AI acoustic analysis and cites a 6.4 million litre per day reduction at an English water utility[5]. An industrial metering startup meters water at Indian factories and campuses[6]. These firms sell equipment or metering.
The gap is an Indian crew that delivers verified leaks as a service to utilities that lack trained staff.
| Difficulty | Hard (53/100) |
| Starting capital | ₹1.2–2.4 Cr |
| Time to launch | 6–9 months |
The loggers can be bought. The difficult parts are government sales, slow payment and training crews to tell a leak from a pump, a transformer or traffic noise on mixed pipe materials. You also depend on the utility to dig, since a leak counts only once it's confirmed.
No sector licence for leak surveys, but selling to cities brings its own paperwork.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't treat a model's score as a confirmed leak. A technician with a ground microphone has to pinpoint it before anyone cuts a road.
Want to work out the AI layer for Leakfinder? Talk to Yash →
Here's what can go south -
I'd pick this for a founder with water-utility or pipeline-maintenance experience who can live with long government sales cycles. If you're expecting software margins then look elsewhere.
Start with two or three private campuses to build case studies, then use an AMRUT start-up pilot to get into one city zone. Prove that at least 70% of flagged spots are real leaks when dug.
Plan for roughly ₹1.2–2.4 Cr for loggers, trained crews and the working capital to survive slow government payments.
No sector licence. You need WPC approval for radio loggers, company and GST registration, and tender or GeM registration to bid for city work.
Yes, any utility with pressurised pipes loses water to hidden leaks. Check local radio rules and how your utilities buy services.
The free Leakfinder validation kit has the line-by-line capital breakdown, the effort scoring, the radio and procurement checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for utilities and campus managers, and 5 assumptions to test. Download the kit →
Thinking about building Leakfinder or something close to it? The make-or-break work is winning a first city zone and getting paid on time once leaks are dug, and I help founders plan exactly that kind of slow sale. 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.