AI is Great

 Fed up with traffic chaos, Bengaluru techie builds AI-powered helmet that reports violators to police instantly

For Pankaj Tanwar, a breakthrough came after discovering that the Bengaluru traffic police accept citizen-reported violations through their app, email, and WhatsApp

AI models analyse each frame to detect four common violations: riding without helmets, jumping dividers, running red lights, and wrong-side driving. (Credit: X/@the2ndfloorguy)

Fed up with errant bike riders and drivers, a 27-year-old software engineer has transformed his helmet into an AI-powered surveillance device that detects violations and reports them directly to traffic police—complete with photographic evidence, video clips, and GPS coordinates.

Speaking to indianexpress.com, Pankaj Tanwar, who is also a passionate bike rider, said he was tired of people breaking laws on the road.

“I have been riding bikes, both for short and long distances. What really annoyed me was how some of the errant riders not only risk their lives but also put other lives at risk on the road. There was no point in arguing or advising anyone on the streets,” he said.

For Tanwar, who is into artificial intelligence at work, a breakthrough came after discovering that the Bengaluru traffic police accept citizen-reported violations through their app, email, and WhatsApp—provided complaints include supporting evidence like photographs, videos, and location details. That became the blueprint for his helmet-mounted enforcement system.

The setup comprises a Logitech webcam attached to his helmet, connected to a Raspberry Pi—a compact computer that processes live video. AI models analyse each frame to detect four common violations: riding without helmets, jumping dividers, running red lights, and wrong-side driving.

Dual-verification to minimise errors

To minimise false positives, Tanwar incorporated a dual-verification system. Detected violations are cross-checked using multiple AI models, including OpenAI’s GPT and Google’s Gemini, before validation. “The system is designed in such a way that it checks frame by frame for nearly five second then if the AI is confident, the report directly goes to the police,” he said

Once a violation is verified, an email containing still images, short video clips, the vehicle’s number plate, and GPS coordinates is dispatched to the police, who then issue challans without further input from Tanwar.

Built primarily from spare parts, the device cost under Rs 4,000. With mass production, Tanwar estimates that the cost could drop to Rs 2,000. Powered by a standard power bank, it runs six to seven hours and is adaptable for cars, taxis, or trucks. 

Tanwar took the outcome of the AI tool to X. He said, “i was tired of stupid people on road so i hacked my helmet into a traffic police device…while i ride, ai agent runs in near real time, flags violations, and proof with location & no plate goes straight to police. blr people – so now ride safe… or regret it.”

This caught the attention of the Bengaluru traffic police, who have invited Tanwar to a meeting. “I am going to meet the Bengaluru traffic police soon to discuss the same issue,” he added.

Tanwar, who has more than 5.5 years of professional experience, has to his credit some projects where technology provides effective solutions to various problems.

One of his projects recently went viral: he connected his stock portfolio with a bedroom light. On X, he posted, “My zerodha stock portfolio now controls my bedroom lights. It turns red when I lose money.”