{"resourceId":"wiplat-sri-aman-ai-acoustic-leak-detection-pilot-2026","versions":[{"version":"external-569ba2d315d9e75955a474705976ad615905ea9d4f88662500da043333ca1ef3","resource":{"id":"wiplat-sri-aman-ai-acoustic-leak-detection-pilot-2026","title":"K-water export pilot in Sri Aman, Malaysia: AI-screened acoustic survey confirmed 23 leaks in four days, vendor reports targets met","organization":"WI.Plat (vendor) for JBALB Sarawak under K-water Export-Led Pilot Procurement","sector":"Water distribution / non-revenue water","geography":"Sri Aman, Sarawak, Malaysia (about 60,000 people); limited transferability to U.S. systems","publishedAt":"2026-11-06","publicationDate":"2026-11-06","eventDate":null,"sourceName":"WI.Plat blog","sourceLabel":"Vendor case study","sourceUrl":"https://wiplat.com/blog/malaysia-sri-aman-ai-leak-detection-pilot/","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["infrastructure","data-security","operating-model"],"finding":"WI.Plat reports a Sep–Oct 2026 pilot for JBALB Sarawak with 75 smart sensors and handheld acoustic devices. 4,004 leak-sound records were collected; AI flagged suspected leak areas for a focused second survey. In four days the team confirmed 23 leaks, with repairs at 16 points checked on site. Reported targets met: leak detection accuracy 95%+ on 2,063 held-out records, 97.96% data collection rate, and 100% field match for AI-suspected points.","sledRelevance":"A government-backed export pilot with stated targets and a held-out test set is a more structured template than typical vendor claims; useful for U.S. utilities designing AI leak-detection pilots with acceptance targets.","evidence":"Vendor case study dated 6 Nov 2026. Accuracy measured on 2,063 records (30% of 6,876 from 50 sensors); field-match metric counts confirmed AI-suspected points but does not report missed leaks. No water-loss volumes or costs reported.","architectureImplications":"Combines fixed sensors, handheld acoustic survey and a cloud platform; AI narrows survey area before human pinpointing.","governanceImplications":"Pilot plan with explicit targets is a good governance pattern; independent verification is absent.","securityPrivacyImplications":"Remote sensor networks and cloud platforms need device and data security; not addressed.","caveats":"Vendor-authored; 100% field match says nothing about false negatives; Malaysian network conditions differ from U.S. systems.","streamIds":["utility-management"],"roles":{"sales":"Sales framing: non-revenue water managers want proof before buying AI leak detection. This pilot is useful mainly as a template: explicit targets for accuracy, data collection and field confirmation, tested on held-out data. Propose a U.S. pilot built the same way. Don't quote the 100% field match as accuracy; it ignores missed leaks and comes from the vendor.","engineering":"Engineering framing: structure pilots with a held-out test set, independent field verification and metrics for both false positives and false negatives. Combine fixed sensors for continuous monitoring with handheld surveys guided by AI. Check data-collection reliability early, since this pilot reported about 98%. Validate against district water-balance changes over several months, not only the count of leaks found.","delivery":"Delivery framing: define targets up front, run a district-metered-area pilot for a few months, and compare AI-guided surveys with conventional sweeps. The distribution or NRW manager owns results. Proposed acceptance criteria are verified leaks per survey-day, measured missed-leak rate and water-loss change in the pilot area. Risks include vendor-defined metrics, seasonal demand effects and short pilot windows that miss slow leaks."},"roleDetails":{"sales":{"customerProblem":"Hidden leaks drive non-revenue water and blanket surveys are slow.","stakeholders":["Distribution manager","NRW lead","Field crews","Finance"],"discoveryQuestions":["What is current NRW and survey cost?","Do you have district metered areas?","How do you verify vendor leak calls?"],"valueHypothesis":"AI-prioritised surveys may find more leaks per crew-day; must be verified locally.","potentialEngagement":"DMA-based AI leak-detection pilot with predefined targets and independent verification.","unsupportedClaims":["Water-loss reduction volumes","False-negative performance","U.S. transferability"]},"engineering":{"fit":"Utilities with leak-survey programmes and DMA metering.","architectureIntegration":"Acoustic sensors and handheld devices feed a cloud platform; outputs go to work orders and GIS.","prerequisites":["DMA flows","GIS pipe data","Field crew capacity"],"constraints":["Vendor data only","Malaysian conditions","Short pilot"],"security":"Secure sensor devices and cloud access; control data sharing.","proofOfValue":["Held-out accuracy","Missed-leak sampling","Water balance change"]},"delivery":{"work":"Select DMA, deploy sensors, run AI-guided and conventional surveys, compare results.","dependencies":["Vendor","Crews","Metering data"],"ownership":"Distribution/NRW manager.","skillsAdoption":"Train crews on AI-guided survey workflow.","governanceCheckpoints":["Target sign-off","Mid-pilot review","Scale decision"],"acceptanceCriteria":["Proposed: verified leaks per crew-day vs baseline","Proposed: missed-leak rate measured","Proposed: DMA loss change reported"],"risks":["Vendor-defined metrics","Seasonal effects","Overgeneralising"]}},"retrievedAt":"2026-11-09T22:08:19.044793-06:00","enrichedAt":"2026-11-09T22:13:59.679686-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Field crews still pinpoint and repair; AI changes where they look.","procurementImplications":"Shows how to write measurable pilot targets (accuracy, data rate, field match) into procurement.","operatingModelImplications":"Moves leak survey from blanket sweeps to AI-prioritised focused surveys.","peerReview":"unknown","sourceVerification":{"openedUrl":"https://wiplat.com/blog/malaysia-sri-aman-ai-leak-detection-pilot/","referenceExcerpt":"In four days, the team confirmed 23 leaks.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}