unthinkable
Case study · A US parking lot management company (unnamed in the source) · 15,000+ lots in 150+ cities

Parking staff read a car's license plate with a phone camera, instead of costly fixed hardware.

The operator, with parking for aviation, healthcare, hospitality and government clients in over 150 US cities, wanted visitors to park and pay quickly and its staff to manage permits and access without expensive plate-reading hardware.

United StatesPublic transit & mobility (live tracking, ticketing, ride-sharing)

What changed

80,000+vehicles scanned, with 99.98% accuracy (as reported)
1000+parking lots using the app (as reported)
67%higher efficiency (as reported)
What stood in the way

What stood in the way.

What stood in the way
  • Fixed cameras and plate readers in parking lots were expensive and made errors.
  • Vehicles move, plates vary in pattern and one frame can hold several cars, which made reading plates accurately hard.
What changed
  • 80,000+vehicles scanned, with 99.98% accuracy (as reported)
  • 1000+parking lots using the app (as reported)
  • 67%higher efficiency (as reported)
What we built

What we built.

  • A native iOS app that detects and reads the plates of several vehicles through the phone camera and sends them to the parking system.
  • A MobileNet V2 machine learning model for on-device detection, trained on 10,000 images of vehicles in day and night, at different distances and in different weather.
  • A confidence check on every frame, keeping only frames scoring above 50%, then Google OCR to read the characters.
  • A 1x-8x zoom slider for distant vehicles, reading cars moving at 50-60 miles per hour.
  • Digital payment so visitors save time at the lot.
  • iOS (native)
  • MobileNet V2
  • Computer vision
  • Google OCR
  • Python
What it shows

What this work shows about us.

Software in place of hardware

A phone app replaced fixed plate readers, so a new lot is equipped in the time it takes to install an app.

We test against the real world

The model was trained on night, distance and weather conditions, and only confident readings reach the parking system.

Want to build similar license plate recognition for parking? Tell us what you want to change. Our AI assistant, Ask Unthinkable, helps you define your requirements, then connects you with the right team.

Start your plan ↗
If we built it today

If we built it today, with AI.

What we would add now, moment by moment.

Booking

An agent would let a driver reserve a space or buy a permit in a chat, check the plate against the booking on arrival and confirm entry. Prices and permit rules stay in the parking system.

Tracking

An agent would warn a driver before their paid time runs out and offer to extend it, and tell staff about cars overstaying. Staff still decide every penalty.

Ask Unthinkable

Want this for your business?

Tell us what you run or what you're building. Our AI assistant, Ask Unthinkable, helps you define your requirements, then connects you with a senior AI engineer.

Send us a note instead

Required

Thanks. A senior AI engineer will reply to you.
u
Ask UnthinkableClarity first, then the right team
Your conversation

An AI agent. This prototype uses prepared answers and sends nothing.