Tome, Trek Collaborate on Using A.I. to Avoid Bike-Vehicle Collisions
After Jake Sigal’s first startup, Livio, was acquired by Ford in 2013, he knew he wanted to stay in the realm of connected cars with his next venture.
At the time of acquisition, Sigal’s company was developing software to enable in-vehicle connectivity, the infotainment-heavy roots of today’s autonomous vehicle innovations. Back then, he saw that a major technological shift was taking place in the industry as self-driving cars zoomed to the head of the future products queue, and he felt that auto manufacturers were not fully prepared for what was to come.
In an Oct. 2013 interview, Sigal told Xconomy he hoped to one day be a part of creating the industry standards that would govern the development of mobility technologies. Four years later, he’s doing just that with a new Royal Oak, MI-based company called Tome Software. This week, the company begins a 12-month collaboration with Waterloo, WI-based Trek Bicycle to develop collision-avoidance systems that use artificial intelligence to reduce the number of bike-vehicle accidents.
The proof-of-concept collaboration will take place at Mcity’s TechLab, an incubator for early-stage mobility startups on the University of Michigan campus. Sigal estimates the total project investment will come to roughly $1.5 million.
Sigal co-founded Tome with Massimo Baldini in 2014 to further delve into software development across a variety of sectors. “We’ve been heads-down for a while,” he explains. “We help large companies connect things that move. Our customers come to us to help figure out the services their customers need.”
Much of Tome’s work, he adds, involves finding the right approach and how to incorporate the latest technologies, such as machine learning, to improve its customers’ products.
Sigal is also an avid cyclist; he and his wife race mountain bikes as part of a local team. “I’ve been involved with bikes my whole life,” Sigal says. “My passions are music—which led to Livio—and biking. I don’t stray too far from my passions.”
Because he’s a bike guy with a background in software and connected cars, he figured he was in a unique position to tackle the problem of how to make bikers more visible to cars—even those without human drivers—in order to avoid collisions.
But he had a theory. If software was created solely to alert drivers to the presence of bikers on the roadway, it would quickly become irrelevant.
“You’d get dings every five seconds, and then you’d eventually turn the notifications off,” he predicts. “What drivers really want to know is if a bike is in a dangerous or vulnerable location. But then how do we define the most dangerous parts of the city?”
That’s a complex thing to determine, he says, and going by just the number of accidents in any given spot is not enough. Drivers also need to take other factors into account, such as weather, geography, and the condition of the road itself.
“We wanted to find the appropriate balance, and to focus first on life-threatening situations,” Sigal says. “A.I. takes all that data and spits out which roads are the most dangerous.”
In the U.S., a combined 45,000 cyclists were injured or killed in vehicle accidents in 2015. A recent market forecast also suggests artificial intelligence will generate $40 billion in revenue for nearly 30 industries over the next decade, Sigal says.
At TechLab, Tome and Trek will test what works and what doesn’t, paying attention to user experience and what’s best for both bikers and drivers. Sigal first met the bike-makers at Trek during a conference in Canada, where he learned more about the company’s research into biker visibility on the roads. Trek found out more about Sigal’s automotive background and a partnership was born.
“For us, it was a no-brainer,” he says. “It’s the perfect peanut-butter-and-jelly match. The key thing is that it’s a cross-company collaboration. This is not a Tome-created-and-owned situation. It’s not a proprietary solution, we’re making it for the industry.” … Next Page »