NHAI / IHMCL ยท Pan-India
TMCC
IoT toll monitoring platform: live health of every toll plaza nationwide, with automated Loose FASTag case routing to banks.
1500+tolls monitored live
30%less toll downtime
50%faster resolution
20M+transactions a day
10TB+lane video a day
3 minof downtime before an alert fires
- Client
- NHAI / IHMCL
- Domain
- National toll operations
- My role
- Data platform: health pipeline, data lake, analytics, ticketing
01
Context
- Plazas ran software from a dozen different vendors
- Nobody could see the whole network at once
- Failures were found by angry drivers, not alerts
02
Before and after
Equipment failuresSpotted by drivers, hours laterAlert after 3 minutes down, ticket raised automatically
Network viewPlaza by plaza, vendor by vendorOne live dashboard across 1,500+ monitored plazas
Toll downtimeNot measuredDown 30%, about $500K a year recovered
Issue resolutionEscalated after complaints50% faster
Operator compensationManual reconciliationCalculated from rules on schedule, with a full audit trail
03
The problem
- No single viewEach vendor's system was an island. Lane uptime, equipment health and revenue could only be read plaza by plaza, if at all.
- Failures found by complaintsA broken barrier or scanner could stay down for hours. Every hour meant lost toll and a backed-up lane before anyone was alerted.
- Data with nowhere to goMillions of daily transactions were generated and never brought together, so trends, anomalies and forecasts were out of reach.
- Compliance by spreadsheetUptime mandates and scheme payouts to operators depended on manual reports from each plaza: slow, inconsistent and hard to audit.
04
How it works
1Sources
- Health monitors on plaza equipment
- Toll systems from many vendors
- National FASTag transaction feed
- Lane cameras
2Ingest
- Equipment heartbeats with downtime alerts
- One adapter per vendor, one common format
- Streaming and scheduled pipelines
3Store & process
- Cloud toll data lake
- Fast analytics layer for slicing by plaza, lane and time
- Video classification
- Rules engine for operator compensation
4Use
- 24x7 command centre
- Dashboards for each role, plaza to ministry
- Automatic tickets and escalation
- Audit-ready compliance reports
05
What we built
- Live equipment health
- Barriers, readers, lane computers, cameras
- Downtime alert within minutes
- Ticket raised automatically
- One data lake
- Every vendor's data in one format
- Next to the national FASTag feed
- Raw data kept in full
- Fast dashboards
- Slice by plaza, lane or hour in seconds
- A view for each role, plaza to ministry
- Policy to payout
- Government scheme rules encoded once
- Operator compensation calculated on schedule
- Every figure traceable to its trips
05
Key insight
In a multi-vendor network, don't ask every vendor to change. Adapt to each one at the edge, and give everyone one version of the truth.
06
The hard parts
- Many vendors. Adapt to each vendor's system instead of changing it.
- Alerts people trust. Alert on real downtime, not every missed signal.
- Money from rules. Every payout can be recomputed from raw data.
- The trade-off. Ran Druid and custom ETL on AWS instead of a managed warehouse - more to operate, but full control over ingestion from 1,500+ heterogeneous plaza feeds.
07
My part
I designed and built the data side of the platform: the equipment-health pipeline, the toll data lake on AWS, the processing and analytics layer for mixed plaza feeds, and the dashboards and automated ticketing on top. The wider TMCC programme, including video analytics and running the command centre, was delivered with the Auriga IT team.
08
What I took away
- Normalise data at the edge of your platform.
- Tune alerts for trust first, then speed.
- When software computes money, make every number reproducible.
Want the detail behind the numbers - architecture, failure modes, what I’d change? Happy to walk through it.
Talk about this project →