Custom car platform Zoox hails from Foster City, California. Its chief executive, Aicha Evans, spoke in a keynote address about the challenges self-driving cars face and suggested they could have a bright future if overcome. It’s also the first time she’s detailed some of the hardware and software stack for Zoox’s custom car platform, which has gone a long way in convincing investors to put more than $800 million into Zoox’s research and development.
Evans believes that building driverless cars in the future will require three core capabilities: artificial intelligence, fully autonomous driving, and a battery-powered electrical framework designed for artificial intelligence. “The new era of mobility requires a radically different approach, and instead of an incremental approach, we need an approach optimized for the human driver,” Evans said.
She did not disclose Zoox’s specific self-driving car design products, but reiterated that it uses a combination of RGB cameras and lidar (sensors that measure the distance of objects by illuminating laser light and measuring the reflected pulses). Each vision sensor can see 270 degrees, and if one of the sensors fails, the car can still maintain a 360-degree view of the environment.
It’s not your average driverless electric car, nor is it your average zero-emission driverless car. Zoox’s model car has four steering wheels, which Evans claims allows it to track trajectories with greater precision than cars with two steering wheels. The self-driving electric vehicle also has a dual powertrain and dual-battery setup, with a combined capacity greater than any single-battery self-driving car today, Evans said.
The aim of this design is to reduce congestion through fleet management and minimize the number of trips back to the base station for overnight charging. Zoox’s new space shuttle-like vehicle is completely driverless and designed to operate in a shared fleet to maximize efficiency and minimize ride time.
This is no different from the self-driving sharing network described by Tesla during Autonomy Day earlier this year. The latter officially released Autopilot hardware 3.0 in early April, which is equipped with a new technology 100% developed by Tesla itself. Autopilot chip FSD.
“When you’re not using it, someone else is, and that’s a way to make better use of resources,” Evans said. “We believe this technology can solve the challenges that cities around the world are currently facing, and with it, we can imagine a world where you can choose not to own or operate a car to carry out the convenience of daily life.”
Like other self-driving vehicles, Zoox uses machine learning algorithms to deal with harsh environments they have never seen before, such as construction zones. They take in visual data and map new paths around obstacles and obstacles, all without human intervention.
Zoox is testing most of the brand’s vehicles in San Francisco and some in Las Vegas, much like Cruise, a vehicle owned by rival General Motors. It’s a conscious choice: As Evans points out, San Francisco’s roads can also be a challenge for human drivers. “Our goal is to develop this technology safer than humans. There are plenty of challenges that can teach us how to be better and safer,” she said.
Perhaps unsurprisingly, Zoox is running simulations using data collected from the real world to continually improve the performance of its systems. The company’s cars drive through city blocks to create topologies, which its engineers use to create three-dimensional representations that AI agents can continually traverse and improve themselves.
Ultimately, when Zoox’s fleet is deployed commercially, Evans believes it will save drivers valuable time. Perhaps more importantly, give them control over their time. She pointed out that it is estimated that human beings spend 400 billion hours driving each year. Drivers in San Francisco alone spend a total of 400,000 hours commuting to and from work every day. The time wasted on the road can be used to do something more meaningful things.
“We’re more productive in a more comfortable environment where we can socialize with friends and enjoy music. Of course, some of us may use those hours to work as well,” Evans said.
Evans asserts that this isn’t the only way Zoox self-driving cars could change the world. They could also reduce the need for parking spaces and buildings, minimize urban congestion (a third of which is caused by drivers looking for parking spaces), and reduce air pollution by about a quarter. In fact, on that last point, Evans said, a fully electric driverless car-sharing fleet could save 85,000 tons of CO2.
In addition, they can also reduce the occurrence of traffic accidents. Statistics show that about 94% of car accidents are caused by human error. In 2016, the top three causes of traffic deaths were distracted driving, drunk driving, and speeding, and the National Safety Council puts an American’s odds of dying in a car crash at 1 in 114.
“We believe that designing self-driving cars from scratch, rather than designing [modified] cars for human drivers, is the right strategy to start a new era,” Evans said. “But we think it’s worth it.”