A video produced by Zoox, an Amazon-owned self-driving startup, shows how driverless technology has advanced to allow an autonomous vehicle to navigate through Las Vegas traffic.
The video shows the drive through the car’s exterior cameras, as well through the software itself that identifies the traffic and plots the best course of action for the vehicle. In the video, the self-driving car merges across three lanes, between vehicles and in close proximity to other cars.
The animated depiction of the driving scene shows how the car detects hazards and roadblocks, as well as a safe following distance behind other cars on the road. Zoox uses a system of LIDAR and radar sensors, as well as four cameras that are mounted on all exterior corners of the vehicle. It says the system helps eliminate blind spots and gives a 270-degree view from each camera, reports Business Insider.
It’s not the first time that Zoox has tested out its autonomous software in Las Vegas. Last year, the company ran an hour-long drive through the city to test out the software in as many complicated scenarios as possible via busy roads, as well as through dense areas where the car would learn to navigate around pedestrians and construction zones.
Zoox has been permitted to test its autonomous driving software with safety operators on public streets since 2016 and has been actively testing its product in Las Vegas, San Francisco, and Foster City.
The startup plans to offer driverless all-electric robotaxis in urban areas and just unveiled its first electric robotaxi in December. Amazon bought the company in 2020 for a reported $1.2 billion and Zoox has yet to reveal when it will launch its ride-hailing service. .
The startup is one of many racing to bring driverless cars to the public. Tesla has been long working on its beta version of its “Full Self-driving” software, while Alphabet-owned Waymo has already begun offering a self-driving ride service in select cities. Though, user videos of Waymo’s autonomous service, as well as Tesla’s beta software, have shown several glitches in the programs — highlighting the many roadblocks self-driving companies face when it comes to developing a program that can drive seamlessly in unpredictable driving scenarios.