When there isn't a global pandemic, nearly 7 million people drive around Houston, and an estimated 77,000 people commute for more than an hour and a half to work. Drivers spend $1,376 and waste 31 gallons of fuel a year — to sit in traffic for what adds up to 75 hours each year.
When Wilson Pulling moved to the city two summers ago, he set out to fix all that traffic-sitting using autonomously driven cars, but not the high-priced ones that Uber and Tesla have designed. These are your regular, three- or four-year-old Honda Civics and Kia Sorentos — the cars you already own.
In 2016, Pulling founded had Aatonomy with his partner, Yang Hu, based on their thesis work from Cornell's Computer Science program. Moving the company south after two years operating out of San Francisco, they aimed not to build the self-driving car of the future, but to make the cars that Houstonians are wading through congested freeways in today drive themselves.
"Everyone doesn't get to buy a Tesla. They're driving their Corollas," Pulling says. "The way autonomy is going right now, that person is never going to benefit. We are the only way."
The company's technology attaches a wireless receiver to the car, which has to be from at least 2016 to work with them. Then, Aatonomy places sensors all along the roads and streetlights. The sensors and receiver communicate with each other, and enable autonomous driving.
Imagine, Pulling says, a 30-mile of I-45 with Aatonomy's sensors. You'd roar up the freeway, handling the controls. Then, the car's computer, under guidance from Aatonomy's network of sensors, would take over. You'd sit back, the car will navigate the traffic along with the other cars — and if all the cars are autonomous, Pulling says, the algorithm could slash congestion. When your car exits the freeway, you'd take back control.
That stretch of freeway would cost $26 million for 200,000 commuters across Houston, Pulling says, but other self-driving cars cost around $250,000 per vehicle — summing up to $50 billion for those same commuters. And Pulling says the Aatonomy system is a safer bet than the way Uber's autonomous driving. Uber's car once killed a pedestrian because, somehow, the company didn't program it to avoid people jaywalking. But because Aatonomy will manage sensors all over the street, the company will be able to monitor potential accidents more quickly than an Uber car would.
"This is a really radically different approach to a technology that, frankly, a lot of people have lost a lot of faith in," Pulling says.
Aatonomy's approach requires a smart city commitment — but the city of Houston is already buying in. First, Aatonomy, a member of the Ion Smart and Resilient Cities accelerator's inaugural cohort, got a short-term project with Aatonomy and Verizon to mount intersection cameras for studying how to prevent collisions with pedestrians on the Northside.
Additionally, the city has also greenlit a two-year pilot with Aatonomy to automate a bus route in downtown Houston. The aim, Pulling says, is making a "proof-of-concept" before rolling out sensors across I-45 — but it's also to use Houston as proof that autonomous driving can be achieved, but from a different angle than Uber.
"Self driving cars don't work. That's our thesis," Pulling says. "That's why we're building self-driving cities."
Ion Accelerator Demo Day
F. Carter Smith
Progress Report helps idenitify learning gaps and common roadblocks for each student. Image courtesy Progress Report