A field worker points the phone camera at a water, electric, or gas meter. SnapMeter reads the digits, flags impossible numbers, and turns them into tenant line items. No smart-meter swap. No IoT box. No app to install.
If your team reads more than a hundred sub-meters a month, you've felt at least one of these. SnapMeter is built for all three.
A team walks every floor with a clipboard or a printed route sheet. They write digits down. Someone keys them into Excel later. The cycle eats half a working week — every month.
A digit gets transposed on the clipboard. Or in the spreadsheet. Or in the billing system. The tenant disputes the invoice. There's no photo. The dispute drags on for weeks.
Field round Monday. Excel reconciliation Wednesday. Billing team imports it Friday. Approvals next week. By the time the invoice reaches the tenant, it's three weeks old and the next read is already due.
The cards marked ★ ONLY SNAPMETER are features that, to our knowledge, no other photo-based meter-reading product on the market currently ships together. Most competitors are either an OCR SDK with no billing, a billing platform that needs smart-meter hardware, or a paper-replacement app where staff still type the digits in.
One product covers the whole cycle: scan the QR, capture the meter, OCR the digits, flag anomalies, calculate the charge, and push the line item to the tenant invoice. Genea and utiliVisor stop at the photo. Anyline stops at the digits. We close the loop.
The meters you have today work. Mechanical dial, LCD digital, water, electric, or gas. The phone is the reader. No electricians. No hardware procurement. No multi-year AMI rollout. Itron and Sensus need to swap your meters; we don't.
Vision AI trained to read both Arabic and Thai numerals on dial, LCD, and digital meters. Returns the reading, the raw OCR string, a confidence score, and notes on what it saw. The reader confirms or retakes — no manual typing on a clean meter.
High consumption (more than 30% jump). Zero. Negative. Excessive (more than 3x average). Stale reading. Low OCR confidence. Each one routes to the supervisor's queue with the photo and last six months of history. The bad reading never reaches the tenant.
Standard Thai utility rates pre-loaded — Provincial Electricity Authority, Metropolitan Electricity Authority, PWA water — with transformer-loss allowance, peak/off-peak, minimum charge, and common-area allocation. You configure once per lease, not per reading.
Basements and meter rooms have no signal. Readings, photos, and confirmations queue locally on the phone and batch-sync the moment the reader is back in range. Routes don't break when the floor doesn't have Wi-Fi.
Every meter gets a printable QR sticker linked to the tenant. Scan to identify, capture to read. Routes are ordered the way the reader actually walks the floor — not the order assets were imported.
The whole reader workflow runs in the phone browser — camera, scan, OCR, confirm, sync. Add to home screen and it launches like an app. Nothing for IT to push, no App Store review, no native build to maintain.
CSV upload for the initial registry — 10,000+ meters with mall, floor, zone, tenant, type, and starting reading.
Every reading kept by tenant, lease, and meter. Twelve-month chart on every tenant page for dispute resolution.
When the meter is damaged or the OCR can't read, the reader types the value and adds a note. The override is logged and flagged.
Decommission the old meter, link the new one to the same QR/tenant, carry the lease forward. The history stays attached.
Approved batches push line items to your existing billing or property-management system via API. Tenant app, accounting, or both.
Four roles, each scoped to the property they belong to. Readers see only their route. Supervisors see only their site.
Every box below is a real screen in SnapMeter. Every arrow is a real action your team takes. Find the workflow closest to yours — that's how you'll use it on day one.
Organisations like: Central Pattana, Siam Piwat, The Mall Group — operators with dozens of malls, thousands of tenants, and a sub-meter on every shop.
Walks the route. Scans the QR on each meter to load the tenant.
Points the phone at the meter face. AI reads the digits.
Reviews flagged readings. Photo + history shown side by side.
Runs the period batch. Reviews each line. Pushes to tenant app.
Sees the bill in their tenant app. Photo proof one tap away.
Organisations like: Sansiri property-management arm, Ananda's MyAnda PMS, Plus Property — residential complexes with hundreds of units, water and common-area electric sub-meters per unit, monthly billing rounds.
Reads water and electric in the same round. One photo per meter.
Reviews flagged units. Approves the period before billing runs.
Pulls the unit-by-unit billing CSV. Imports into the existing PMS.
Receives the bill with photo evidence. Disputes drop to near zero.
Organisations like: WHA Industrial Estate, Amata, Hemaraj — operators leasing serviced factory units where each tenant pays MEA/PEA-rate utilities by sub-meter.
Reads the high-voltage sub-meter per factory unit. Captures CT ratio.
Spots a 3x jump in factory unit 14. Requests a re-read.
Applies PEA TOU rate per lease. Generates the tenant invoice line.
Sees its monthly utility cost broken out by factory unit.
Organisations like: Provincial Waterworks Authority (PWA), Metropolitan Waterworks Authority (MWA), or a regional electric branch — pre-AMI areas where door-to-door reading is still the norm.
Walks the assigned route. Phone replaces the clipboard.
Photographs the dial. AI reads ๐–๙ Thai numerals on older meters.
Reviews skipped houses, dog-on-property notes, and anomalies.
Receives the period file. Bills go out the same week, not three weeks later.
SnapMeter is built first for Thai property operators, condo juristics, and provincial utility branches — and then for everyone else. The list below is what you would otherwise spend three months adding to a generic OCR meter app.
Older water and electric dials still print Thai numerals. The vision model reads both ๐–๙ and 0–9 — and we test on both.
Provincial and Metropolitan Electricity Authority residential, business, and industrial rates pre-loaded. Time-of-use, transformer loss, peak/off-peak, minimum charge.
Provincial and Metropolitan Waterworks Authority cubic-metre rate tiers for residential and commercial customers, applied automatically per lease.
Readers work in Thai. Supervisors and admins toggle to English in one click. Same data, same reading, two languages.
Default Singapore-region. Thai-region managed-tenant deployment available for PDPA-bound deployments — specify residency at scoping.
Common-area electricity and water split across units by configurable formula — area, headcount, or fixed share. Calculated, not eyeballed.
Default 24-month retention on every meter photo. Tenant disputes resolved by pulling the original capture, not a regenerated screenshot.
Line items match the Revenue Code format your finance team expects — VAT, debit/credit notes, period code — when pushing to your billing system.
Here's what changes in the first billing cycle for a typical property with 1,500–3,000 sub-meters across multiple sites.
Per-meter capture time. Scan the QR, photograph the meter, confirm the reading. Manual transcription drops out of the loop.
Reading round to tenant invoice. The same week the meter is read, the line item is on the bill — no Excel reconciliation in between.
Wrong digits keyed from clipboard to spreadsheet to invoice. Photo-to-OCR removes most of the typing — and the OCR confidence is logged.
Meters needing a second walk because the first reading didn't reconcile. Anomaly checks at capture catch most of them before the round ends.
In operational terms, a property team running a 2,000-meter monthly cycle gets back roughly four reader-days per round — every round. In financial terms, billing dispute volume drops because every line item has photo evidence one tap away. In compliance terms, every reading carries the timestamp, GPS, capturer ID, and the original photo, retained for the period your auditor or PDPA review expects.
Figures are derived from the SnapMeter reference build's product specification, pilot estimates, and our founder's prior enterprise delivery of field-data and OCR-driven systems across Southeast Asia (2017–2026). Individual results vary by meter mix, site count, and existing process maturity.
Most teams are running their first real billing cycle on SnapMeter by week 8. Here's what each phase looks like at one site of around 1,500 meters — multi-site rollouts compound onto this baseline.
Meter list imported, QR labels printed and stuck. Tenants and leases loaded. Rate tables configured per lease. Routes ordered the way readers actually walk.
Readers using SnapMeter for the daily round — scan, capture, OCR, confirm. We're on hand for OCR tuning, route tweaks, and anomaly thresholds based on what your team finds.
Run a real billing cycle at the pilot site. Calculate, approve, push to your billing or tenant app. Finance reviews the output to confirm it matches what they used to produce manually.
Your team operates SnapMeter day-to-day without daily support. We monitor and respond to issues in the background, and start the conversation about additional sites.
Investment is sized to your portfolio, the number of sites, and the meters in scope — discussed in person or on a call when we meet, not on a public price list. Property operators and regulated utility branches usually have specific procurement processes; we'll work within yours.
No. The QR scan, camera capture, OCR result, and confirmation flow all run in the phone's web browser. There's nothing to download from the App Store and nothing for IT to push to staff phones. "Add to Home Screen" lets it launch like an app, but it's still the browser underneath.
Yes. Photos, scans, and reading values queue locally on the phone when the network drops. The whole offline round syncs the moment the reader is back in range — within about 30 seconds. The supervisor doesn't see the round as "missing"; they see it as "syncing."
No. The whole point is the meters you already have keep working — mechanical dial, LCD digital, water, electric, gas. The phone is the reader. This is the wedge against AMI vendors like Itron and Sensus, who need a multi-year hardware rollout to deliver a similar outcome.
Each OCR result comes back with a confidence score and any notes the model flagged ("display partially obscured," "glare on dial," etc.). Below a configurable confidence threshold, the reader is prompted to retake. If the meter face is genuinely unreadable, the reader types the value, adds a one-line note, and the override is logged for the supervisor.
Yes. The reference build pushes line items by API into a tenant-app (CPN's SERVE was the first integration target). For other property managers and utilities we add the integration to your existing billing, ERP, or PMS during pilot scoping. CSV export is always available as a fallback.
By default, the hosted version runs on Singapore-region infrastructure for cost reasons. For PDPA-bound property operators and regulated utility deployments, we offer a Thai-region managed-tenant option — specify residency at scoping. We do not move data between regions without your written consent.
SnapMeter is built by Inline One Systems, a Bangkok-based product studio. Our founder previously built and shipped enterprise field-operations and OCR-driven systems across Southeast Asia between 2017 and 2026. The reference build of SnapMeter was developed against a Thai multi-mall property operator's scope, and we carry that operational pattern forward into the productised version.
A 30-minute walk-through. Bring a sample of your current meter list, your tariff or rate sheet, and one site's worth of routes. We'll demo SnapMeter with your real meters — capture, OCR, anomaly, calc, push. If the shape fits, we scope a structured engagement at one site. If it doesn't, you leave with a clearer view of what your operation actually needs.