Technology | Can AI put an end to car accidents?

Every year, around 1.35 million people are killed in crashes on the world’s roads, and 50 million more are seriously injured, according to the World Health Organization. In the United States, the death toll has increased significantly during the pandemic, resulting in the largest increase on record over six months, according to estimates from the United States Department of Transportation. Speeding, distraction, drunk driving and not wearing a seat belt are the main causes.

Artificial intelligence (AI) is already being used to improve driving safety: mobile phone apps that monitor driving behavior and reward safe drivers with perks and connected vehicles that communicate with each other and with road infrastructure .

But what awaits us? Can AI do what humans can’t? And will the technology develop before the proliferation of self-driving cars?

“In my opinion, there’s too much hype around AI, road safety and self-driving vehicles — it’s very overblown,” said David Ward, president of the Global New Car Assessment Program, an organization not-for-profit based in London. According to him, we must focus on “the fruits at hand and not on a distant utopian promise”.

Defenders like Ward are turning to beneficial, low-cost intermediate technologies available today. A great example is Intelligent Speed ​​Assist, or ISA, which uses artificial intelligence to manage a car’s speed through on-board cameras and maps. This technology will be mandatory in all new vehicles in the European Union from July, but it has yet to take hold in the United States.

“Smart Eyes”

Acusensus, based in Australia, is one of the companies using artificial intelligence to improve road safety. Its cameras — “smart eyes,” as Acusensus calls them — use high-resolution imagery in conjunction with machine learning to detect dangerous driving behaviors that are often hard to notice and enforce.

“We have technology that can save lives,” said Mark Etzbach, the company’s vice president of sales for North America.

The patent-pending technology, which unlike the human eye is unaffected by weather conditions or high speeds, can visualize and record behavior inside the vehicle, Etzbach explained. The cameras can be installed on existing road infrastructure, such as bridges, traffic signs or mobile structures. The images are then optimized for the AI, which is trained at specific parameters.

Acusensus’ algorithms can determine with a high degree of probability whether a particular driver is engaging in risky behavior, the company claims.

Such technology would allow law enforcement to clearly see if a driver is holding something besides the steering wheel – a phone, for example – and if they are looking down to text. (Invisible flash provides clear penetration into the windshield.)

The technology was developed by Acusensus co-founder Alexander Jannink after a friend and fellow software engineer was killed while riding a bicycle in 2013.

“He was struck and killed by an intoxicated driver who is also believed to have been distracted,” Etzbach said.

heads-up

The company’s main product, Heads-Up, was first rolled out in 2019 in New South Wales, Australia. The Heads-Up system captures images which are then scrutinized by authorities to determine the likelihood of a breach. In the first two years, the company says, the state saw a 22% reduction in fatalities and a reduction in phone use of more than 80%. The technology is currently being rolled out in New South Wales and Queensland, with additional pilot projects elsewhere in Australia and overseas.

The next iteration of this technology, Heads-Up Real Time, is slated for deployment in the United States. Data and images would be sent in real time to officers in patrol cars, who could then view them on laptop computers.

Acusensus’ technology can also be used to detect ‘hot spots’, helping to identify where authorities need to improve law enforcement, make changes to infrastructure or pass new legislation. Over the past several months, the company has performed demonstrations and evaluations for a number of local law enforcement agencies and state Departments of Transportation.

During an 18-hour assessment in August of a high-risk corridor in Missouri, which averaged 3.5 crashes a day, more than 11,000 vehicles drove through. At least 60% of drivers were speeding, an average of 6.5% used a mobile phone, more than double the national average, and just under 5.5% engaged in two simultaneous risk behaviors while driving.

“Last year they had one of their record years for traffic fatalities,” Etzbach said. They want to change policy so they can address some of these road safety issues. »

The technology is attracting growing interest in states.

“We have contracts with two states for data projects and are in discussions with many others,” Etzbach said. One such state, Indiana, is “piloting technology for law enforcement assessment.”

Technology similar to that of Acusensus is also being considered in Europe. Alexandre Santacreu, secretary general of the European Metropolitan Transport Authority in Paris, said large-scale data collection has huge potential for use in preventing collisions on road networks.

Flat

Not everyone is convinced of the merits of computer learning. According to David Ward, president of the Global New Car Assessment Program, humans are always better than artificial intelligence.

“An observant driver who makes eye contact with a pedestrian can determine whether that person intends to cross the street or not. AI is not able to do that, not yet,” he said.

As with many AI innovations, the technology also raises privacy concerns.

“It’s that classic question of how much intrusion we want in our lives to keep us safe,” Ward said.

“We share our road space, and there are limits to our freedom in the closed environment of cars, but that involves a greater degree of intrusion,” he said of the surveillance technology. This is what AI brings to us. »