Vision-Based Automotive Safety Systems Go Mainstream


By Jeff Bier
Founder, Embedded Vision Alliance
Co-Founder and President, BDTI
This blog post was originally published at EE Times' Automotive Design Line. It is reprinted here with the permission of EE Times.

Welcome to this new blog on machines that see!

For most of my career, computer vision has been filed away in my mind as one of those cool technologies that was just too complex, expensive, and temperamental for me to use. But times change. Today we find computer vision integrated into a growing range of high-volume products, including game consoles, smartphones, and cars. I use the term "embedded vision" to refer to this proliferation of computer vision into embedded systems, mobile devices, PCs, and the cloud — creating "machines that see and understand."

In this blog, I'll be exploring this exciting field — including applications, enabling technologies, and design challenges. And I'll be inviting expert colleagues to contribute their insights as guest contributors.

Automotive safety is one embedded vision application that is growing fast. Vision-based "automotive driver assistance systems" perform many functions, such as forward collision avoidance, blind spot monitoring, and lane-keeping. While Google's self-driving cars have grabbed the headlines, luxury auto makers have been shipping vision-based driver assistance systems for several years without much fanfare. The fact that established automakers are deploying this technology is significant, because it tells us that the technology is reliable: A company like Daimler isn't going to risk its reputation by shipping a new technology without very extensive testing.

I'm excited about the potential for these systems to reduce the number of accidents and save lives. According to the US Centers for Disease Control, "Motor vehicle crashes are one of the leading causes of death in the U.S. More than 2.3 million adult drivers and passengers were treated in emergency departments as the result of being injured in motor vehicle crashes in 2009." I'm terrified on a daily basis when I see drivers rolling down the highway with their attention on things other than the road.

A couple of years ago, I wanted to try out vision-based safety features for myself, but wasn't prepared to buy a new luxury car. So I installed an aftermarket vision-based safety system (from industry pioneer Mobileye) in my car. You can see a video of me demonstrating the system here. While it's not perfect, I've been pleased with the system's performance, and it has definitely made me a safer driver.

When new automotive safety features are effective, they migrate from luxury models down to less-expensive models. As production volumes grow, these features become less expensive, and eventually they become ubiquitous. Today, features like air bags, anti-lock brakes, and stability control are universal. So it's not surprising that we are now beginning to see vision-based safety systems in mid-market cars like the 2014 Subaru Forester. Read the ExtremeTech review of the Forester here to get a sense of how it works.

With vision-based safety systems showing up in mid-market cars, we'll soon see hundreds of thousands of these systems deployed — and that should start to put a real dent (pun intended) in the number of car accidents.

Here you’ll find a wealth of practical technical insights and expert advice to help you bring AI and visual intelligence into your products without flying blind.



1646 N. California Blvd.,
Suite 360
Walnut Creek, CA 94596 USA

Phone: +1 (925) 954-1411
Scroll to Top