China's Baidu Announces ‘Apollo Air' Which Supports Self-driving Vehicles Using Real-time Data From 5G-Connected Roadside Units

author: Eric Walz   

A Baidu Apollo self-driving taxi.

China's tech giant Baidu is expanding its Apollo autonomous driving platform by adding vehicle-to-everything (V2X) communications technology in which autonomous vehicles communicate with roadside sensors using ultra-fast 5G cellular networks.

The new connected platform is called "Apollo Air" which refers to Baidu's Apollo open autonomous driving platform, which is an open platform designed to foster innovation and speed up the development and rollout of autonomous driving technology through collaboration with industry partners.

The Apollo Air project was unveiled at an academic forum hosted by Institute for AI Industry Research (AIR) on May 13.

The launch of Apollo Air is part of Baidu's partnership with AIR at Tsinghua University. It's one of the most advanced V2X projects to provide reliable safety redundancy for autonomous driving, according to Baidu.

The Institute for AI Industry Research (AIR) was established in 2020 to promote social progress in China using AI-powered technology.

"Tsinghua University has been committed to breakthroughs in core AI technologies, and Apollo Air is an exploration of the V2X field in China," said Dr. Ya-Qin Zhang, Dean of AIR. 

V2X technology that allows vehicles to communicate with their surrounding environment, including traffic signals and signage, 5G connected roadside units (RSUs), buildings, as well as other vehicles. It's instrumental for the successful implementation of fully autonomous vehicles, according to Baidu. 

Apollo Air will be the world's first and only vehicle-to-everything (V2X) technology that enables SAE Level 4 autonomous driving on open public roads that's further supported with data from roadside sensors.

According to Baidu, most autonomous capable vehicles today rely entirely on their embedded perception systems including lidar, radar and cameras for safe navigation. However, these perception systems can be made safer when combined with additional real-time data from roadside units (RSUs). The data from the RSUs also adds another layer of redundancy.

Baidu says this added benefit is one of the reasons why it developed Apollo Air, as it works to commercialize autonomous mobility in China.

How it Works

Apollo Air can improve the perception capabilities of an autonomous vehicle by what Dr Tao calls "vehicle-road-cloud coordination." With Apollo Air, even a regular vehicle without sensor equipment can also achieve some high-level autonomous driving capabilities by receiving data from RSUs.

According to Baidu, challenging problems and edge cases, such as driving in inclement weather, bright sunlight or low light conditions, are yet to be fully resolved. But these autonomous driving challenges can be better met using smart RSUs to intelligently communicate with vehicles.

If an autonomous vehicle is traveling down the road in blinding sunlight, vehicle cameras that identify the color of the traffic signal may not be able to identify the light status. However, if an autonomous vehicle is able to communicate with RSUs at intersections, traffic light status can be fed to the vehicle in real-time to supplement the vehicle's cameras.

Roadside V2X equipment can continuously feed self-driving vehicles real-time information such as the location and speed of surrounding cars or other objects, which will greatly expand the collective field of view and eliminate blind spots. 

Baidu and AIR have tested Apollo Air's capabilities in L4 autonomous driving scenarios at intersections in Beijing, Guangzhou, and Cangzhou. 

In the future, Baidu will gradually integrate parts of Apollo Air's evolving road sensing capabilities in its production-ready V2X offerings and provide reliable road sensing data for robotaxi operation and advanced driver-assistance systems (ADAS).

"Apollo Air enables a high degree of coordination between sensor-less vehicles, smart roads and data clouds through a series of miniature roadside sensors with 5G and V2X wireless communication technology," said Dr. Ji Tao, General Manager of Intelligent Transportation Product Development at Baidu. 

Additional data from RSUs can help eliminate blind spots for autonomous vehicles near intersections.

Apollo Air will also be part of Baidu's open platform. Baidu will release R&D data and use cases of Apollo Air regularly under open source and open standards principles. Baidu's goal is to build Apollo Air into an industry-wide sharing platform.

There's been a push to develop V2X and intelligent connectivity platforms over the past few years in China. The country's Ministry of Industry and Information Technology recently laid out its official positioning for autonomous vehicles using V2X technology. As a result, China is ramping up the construction of V2X infrastructure.

For the past several years, Baidu has been working on advanced self-driving vehicle technology as part of its Apollo open development platform in addition to its core internet related businesses. 

The tech giant aims to be one of the world's first companies to scale and commercialize autonomous driving technology in China, which is the world's largest auto market.

Baidu's recent work in the field includes developing its Apollo Computing Unit (ACU) for self-driving cars, deploying robotaxis, building HD Maps, advanced driver assist systems (ADAS), 5G connected vehicle technology and autonomous buses.  

Since launching Apollo in 2017, the platform has grown to include over 100 industry partners working with Baidu to advance autonomous driving and related technology at scale.

Baidu was the first company in China to pick up passengers in autonomous vehicles when it first opened its "Apollo Go" robotaxi service to the public in April 2020. Riders in China can summon one of the company's Apollo Go robotaxis using Baidu Maps or the Apollo Go smartphone app.

Soon these autonomous Apollo vehicles might be communicating with RSUs and other infrastructure.

Eric Walz

Originally hailing from New Jersey, Eric is a automotive & technology reporter covering the high-tech industry here in Silicon Valley. He has over 15 years of automotive experience and a bachelors degree in computer science. These skills, combined with technical writing and news reporting, allows him to fully understand and identify new and innovative technologies in the auto industry and beyond. He has worked at Uber on self-driving cars and as a technical writer, helping people to understand and work with technology.