Chitti app logo

Chitti is better in the app

Coins, streaks and builds — free on Android & iOS

Get the app
Chitti logoChitti

How We Created a Fake Traffic Jam?

8 Minutes

Fake traffic jam experiment for students: discover how Google Maps detects traffic, crowdsourcing, and machine learning. A Tamil technology lesson.

  • Collecting phones and riding slowly to create fake traffic on maps
  • How Google Maps uses crowdsourced phone data and speed limits
  • Using a picture puzzle to explain machine learning and programming
What this lesson teachesRead the lesson as text

A fake traffic jam can be created on digital maps by moving a large collection of active smartphones very slowly along a street, tricking navigation algorithms into treating them as slow-moving vehicles. Digital navigation services rely on continuous location and speed data transmitted from users to determine road conditions in real time.

How maps detect traffic

  • In the past, digital maps relied on years of historical traffic records, but this data could not predict sudden events like road accidents or ongoing construction.
  • Road sensors and traffic cameras were also used to collect real-time data, but installing them on every road proved difficult and expensive.
  • Modern mapping services use crowdsourcing, gathering live location, direction, and travel speed directly from user smartphones through GPS, Wi-Fi, or cellular towers.
  • When devices move at the speed limit of forty to fifty kilometres per hour, roads show as clear, but speeds dropping below twenty kilometres per hour trigger a heavy traffic warning.
  • Additional live data comes from mapping applications like Waze, where travellers manually report potholes, collisions, and roadwork.

The fake traffic experiment

To test this system, dozens of smartphones were collected into a backpack and carried on a scooter. Riding through multiple streets initially failed to create a traffic warning, because road width and other road conditions are also factored into the algorithm. After hours of riding and pushing the scooter slowly, travelling at a crawling pace through a narrow street convinced the mapping system that vehicles were jammed. The empty street was then marked with a red traffic line on the screen.

How machine learning helps

Processing billions of incoming data points from around the world is impossible for human workers alone. In traditional computer programming, people supply both the data and the rules to produce an answer, similar to following fixed instructions to solve a jigsaw puzzle. Machine learning instead gives the computer the data and the finished results, allowing the machine to learn the rules on its own. Digital maps use machine learning models running on supercomputers to detect patterns, predict traffic flow, and suggest the fastest routes to drivers.

More lessons

Browse all lessons →