How I created Driver on Phone Recognition using AI 🤖

How to detect Drivers on the Phone? I will explain how my proof-of-concept system operates and how it can be applied in-field. The Netherlands is already testing a similar system.

Remco van Akker


Phone use while driving
Geber86 | iStock | Getty Images

This article is a short summary of my paper ”DOPR: Driver On Phone Recognition using Artificial Neural Network”. Not all technical & functional details are explained in this article. For more details, you can read my paper which is available on my LinkedIn

The problem we want to solve

Have you ever noticed the person in front of you (or at your side), while driving, was calling or texting, and you felt unsafe? Have you ever heard of, or witnessed, a car accident caused by mobile phone distraction? These issues can now be prevented by implementing a system in today’s society called driver-on-phone detection using artificial intelligence.

Let’s get the facts straight first.

In the United States, The National Safety Council reports that cell phone use while driving leads to 1.6 million crashes each year.

In 2017 in my home country (The Netherlands), more than 20,000 people were injured, and 629 people lost their lives in a car accident. The Dutch are aware of the dangers of using their phone while taking part in traffic. Yet, mobile phone is often used in traffic. Here are a few statistics from Dutch research by Interpolis.

  • 92% of the Dutch consider using a mobile phone in the car to be dangerous.
  • 86% of co-drivers feel unsafe next to someone using their mobile phone while driving.
  • 80% of the co-drivers say they will say something if this happens.
  • Yet 62% of the Dutch also use the mobile phone behind the wheel.

However, if the police catch you driving while using your phone, you receive a fine. The current fine in The Netherlands for using your phone while driving is 230 euros. This is a painful fine, but it’s not an efficient way to prevent phone use on the road.