Skip to main content

Brain4Cars analyzes driving behavior with AI to predict and prevent road accidents

As cars gets smarter, tech developments in our dashboards aim to help humans keep up, or at the very least to keep human drivers alert behind the wheel. To make this happen, researchers have developed a deep learning neural network called Brain4Cars that combines the many approaches of top-of-the-line smart car technology, and augments them with artificial intelligence.

Cars obviously aren’t the only things getting smarter, and as our gadget collections pile up we’re prone to becoming ever more distracted. Even with existing road safety regulations and legislation to keep drivers safe, 33,000 people die in road accidents every year. What’s more, over 90 percent of road accidents in the U.S. can be attributed to driver error, according to the National Highway Traffic Safety Authority.

Many car manufacturers have already taken to installing safety systems in new models that alert drivers to dangerous behaviors and external environmental risks. Brain4Cars basically takes all these technologies and rolls them into one, combining safety measures like internal and external cameras with GPS, vehicle dynamics reporting, and an extensive database of driver behavior recordings.

What sets Brain4Cars apart is what it does with this comprehensive, real-time data collection. By feeding information from the data sources into a proprietary artificial intelligence, the system is able to reliably predict individual driver behavior in real-time, and anticipate errors up to 3.5 seconds in advance with an 80% accuracy rate.

Researchers from Stanford and Cornell modeled the Brain4Cars system on Recurrent Neural Networks with Long Short-Term Memory. The artificial intelligence analyzes in-car visual cues like the driver’s facial expressions and sightlines, while external camera feeds provide the system with a contextual understanding of the safety and appropriateness of driver choices in relation to the situation unfolding on the road.

In theory, a deep learning architecture like this could come to learn the habits of each individual driver it assists. Over time, Brain4Cars would be able to predict driver accuracy with greater precision and recall as data from the specific driver becomes part of the network’s knowledge base.

This background contextual information provided to the neural network was amassed over the course of 1,180 miles driven collectively by ten different drivers in California. Road areas included freeway and city driving to expand the AI’s contextual knowledge for commonly encountered road situations and external car environments.

Even driverless cars get their fair share of heat when it comes to on-road errors, but prediction about the future of autonomous cars suggest that accidents and collisions will be mostly due to human error — not robotic drivers. But instead of replacing human drivers entirely, Brain4Cars combines the best of driver safety tech to bring a bit of robotic precision support to humans behind the wheel.

Editors' Recommendations

Chloe Olewitz
Chloe is a writer from New York with a passion for technology, travel, and playing devil's advocate. You can find out more…
Mind-reading A.I. analyzes your brain waves to guess what video you’re watching
brain control the user interface of future eeg headset

Neural networks taught to "read minds" in real time

When it comes to things like showing us the right search results at the right time, A.I. can often seem like it’s darn close to being able to read people’s minds. But engineers at Russian robotics research company Neurobotics Lab have shown that artificial intelligence really can be trained to read minds -- and guess what videos users are watching based entirely on their brain waves alone.

Read more
MIT is teaching self-driving cars how to psychoanalyze humans on the road
mit algorithm predict drivers personality car driver behind wheel

In March 2004, the U.S. Defense Advanced Research Projects Agency (DARPA) organized a special Grand Challenge event to test out the promise -- or lack thereof -- of current-generation self-driving cars. Entrants from the world's top A.I. labs competed for a $1 million prize; their custom-built vehicles trying their best to autonomously navigate a 142-mile route through California’s Mojave Desert. It didn’t go well. The “winning” team managed to travel just 7.4 miles in several hours before shuddering to a halt. And catching fire.

A decade-and-a-half, a whole lot has changed. Self-driving cars have successfully driven hundreds of thousands of miles on actual roads. It’s non-controversial to say that humans will almost certainly be safer in a car driven by a robot than they are in one driven by a human. However, while there will eventually be a tipping point when every car on the road is autonomous, there’s also going to be a messy intermediary phase when self-driving cars will have to share the road with human-driven cars. You know who the problem parties are likely to be in this scenario? That’s right: the fleshy, unpredictable, sometimes-cautious, sometimes-prone-to-road-rage humans.

Read more
Will autonomous cars eradicate driving jobs? Don’t hold your breath
tusimple-self-driving-truck

Automation, job security, and truck driving were center stage at last week’s Democratic debate. Andrew Yang and Elizabeth Warren in particular sparred over whether automation was going to kill jobs and wipe out the trucking industry. Without a doubt, the next 10 years will see drastic changes to the way cars and trucks move us and our cargo around.

But the next decade of changes will entail much smaller changes than many people – presidential candidates included – are expecting. The 2020s probably won't bring flying cars, solar roads, or robot gas pumps. Like much of our current world and technology, the 2020’s will most likely see a melding of the amazingly futuristic with the frightfully analog.

Read more