Last week, autonomous car startup Drive.ai announced a self-driving pilot would launch in Frisco, Texas in July. Today, the company gave a behind-the-scenes glimpse at the program’s underlying technology.
A newly published video shows one of Drive.ai’s cars traversing private and public roads, encountering pedestrians, cyclists, and other vehicles that its object recognition system identifies and nimbly routes around.
“[Roundabouts and intersections] often [require] care to negotiate traffic merging from multiple points,” Drive.ai wrote in a Medium post. “As the car proceeds to drive on public roads by crossing a 6-lane intersection, our system looks both ways for a long distance to understand what is happening and how to proceed. Low-angle sunlight can make it difficult to see, [so] our system fuses the inputs from multiple sensors to ensure accurate detection and tracking.”
Drive.ai, which was founded by graduate students out of Stanford University’s Artificial Intelligence Lab, has been working closely with the Frisco Transporation Management Association (TMA) ahead of the launch of its car service. It plans to give over 10,000 people rides in self-driving cars over the course of six months from fixed locations in geofenced areas of the city.
Drive.ai’s service will be the first public self-driving car test in Texas, but not the first in the country. Google-owned autonomous car company Waymo rolled out a fleet of self-driving cars in Phoenix, Arizona this year. And Uber embarked on autonomous car tests in San Francisco, Pittsburgh, Phoenix, and Toronto, programs which it temporarily halted after an accident involving a self-driving car in Arizona.
In May, Drive.ai told VentureBeat that it is working to “educate and engage” members of the Frisco community, including by placing signage throughout the planned routes. The self-driving cars will be painted a “highly visible” orange color and feature displays that communicate their intended actions to drivers and pedestrians on the road.