5G Provides the Missing Puzzle Piece for Self-Driving Cars

  • 5G Provides the Missing Puzzle Piece for Self-Driving Cars

Imagine an autonomous car self-navigating through an intersection by communicating with other vehicles, as well as with traffic signals and sensors that alert it about oncoming pedestrians. It might sound futuristic, but that’s exactly what visitors to Barcelona’s L’Hospitalet de Llobregat district saw during a 5G connected cars pilot project.

Ateca and Arona cars, made by Seat, used Telefonica’s 5G network and Ficosa’s C-V2X (Cellular Vehicle-to-Everything) technology to communicate with other cars, cyclists, and even traffic signals. The traffic lights used thermal cameras to detect pedestrians in the crosswalk, and displayed a warning message in the car’s control panel whenever they saw someone crossing the road. Connected bicycles communicated with cars around them to inform them of their location and warn them of any potential collisions. In cases of low visibility, non-moving cars activated emergency lights and automatically notified other approaching vehicles through a direct communications interface.

Faster lane to autonomy

The autonomous cars tested on the streets of Barcelona sound exciting, but all their engineering, software and high-tech gadgets alone aren’t enough to make self-driving a reality. To take off, they will need the new capabilities of a 5G mobile network’s lightning-fast speed, incredibly low latency and the capacity of carry massive number of connection simultaneously.

“5G multi-gigabit speeds, distributed edge networks, and low-latency services can help these capabilities to become a reality in years to come,” says Chris Penrose, president of IoT solutions at AT&T. “5G will eventually allow for the necessary distribution of processed data to meet and exceed the needs of the automated vehicle.”

Penrose is right. Autonomous cars can’t do simple things, such as detect a pedestrian or swerve around a curb, without data processing capabilities that can enable actions as fast as human reflexes. Even a split-second delay in communicating a potential collision could mean the difference between safe road traffic and a major crash.

5G integrated where the rubber meets the road?

Not surprisingly, a slew of automakers, such as Audi, BMW, Daimler, Ford, Hyundai and Toyota, are integrating 5G technologies into their vehicles.

Take Ford, for instance. It has committed to making sure that all of its new cars in the United States will have 5G-powered CV2X communications capabilities by 2022. “Billions of dollars already are being spent as the cellular industry builds 5G networks, so we think the timing is perfect to give our vehicles some of the natural skills we use every day to get around,” says Don Butler, executive director of Ford Connected Vehicle Platform and Product.

Likewise, Daimler, the maker of the Mercedes-Benz luxury cars, has a plan with its biggest shareholder, Zhejiang Geely Holding Group, to assemble all-electric, 5G-enabled Smart micro-cars in China for export. The first car is expected to be on the road by 2022. And Hyundai Mobis, the parts manufacturing arm of Hyundai Motor Group has partnered with telecommunication company KT to jointly develop technologies for connected vehicles running on the 5G network.

Telecom companies are rolling up their sleeves too. Audi is a founding member of the 5G Automotive Association (5GAA), created to connect the telecom industry and vehicle manufacturers to develop end-to-end solutions for future mobility and transportation services. Audi plans to launch 5G connected cars in the very near future and has teamed with telecom companies including Huawei, who last year launched what it claimed was the world’s first 5G hardware for self-driving cars. The 5G module is an element in their effort to commercialize a 5G network for the automotive sector.

That’s not all. Several telcos are setting up fake cities as test beds for vehicles. Samsung, for instance, has collaborated with the Korea Transportation Safety Authority to build test-bed K-city in Korea to create an environment for autonomous vehicles and connected cars to test 5G in real-life road conditions, crossings and tunnels.

Driven by data infrastructure

All the promising experiments with 5G cars won’t go anywhere until the right infrastructure is in place. What’s more, autonomous vehicles require excellent roads and harmonized government regulations, as well as smart signage, sidewalks, and curbs.

But the most significant challenge is managing the volume and complexity of data that self-driving cars will generate. A single autonomous vehicle is estimated to generate about 4 terabytes of data during an average day.

Much of this data, generated by a myriad of sensors inside self-driving cars, must be processed closer to the action. That’s where the evolving edge computing infrastructure comes into play. This approach allows for real-time data processing both on-board the vehicle and in nearby processing centers that might collate the data from surrounding traffic and traffic control sensors and devices. This is the technology that enables cars to detect objects around them, and make life and death decisions in milliseconds.

More than just crash avoidance

Of course, not all types and quantities of traffic-related data needs to be acted upon immediately — and there’s some limit to the localized computational processing capability and memory that these cars can have on-board. So some data processing that’s less time sensitive can be captured and analyzed in aggregate at local edge data centers, as well as some that continues to be handled by communicating with remote cloud servers.

The onus is on car companies and city planners to capture, manage, transmit, analyze and secure data about every vehicle, every slowdown, every pedestrian and every bump and every pothole. Some smart city developers are already experimenting with machine learning algorithms capable of analyzing data to find better ways to fill potholes more quickly, provide traffic updates, respond to emergencies, and assess infrastructure investment decisions.

According to Brooks Rainwater, director of the Center for City Solutions at the National League of Cities, data generated from self-driving cars will provide cities with “a more granular viewpoint into everything from infrastructure wear-and-tear to detailed traffic flow information and even sidewalk congestion patterns.”

As 5G goes mainstream, autonomous cars will generate large amounts data that will need to be continually managed and analyzed in order to make smart cities a reality. Getting all of this right may be a bumpy road, but the end result may be a new chapter in mobility that rivals the invention of the automobile itself.


About the Author:

John Paulsen
John Paulsen is a "Data for Good" advocate, with more than 20 years in the data storage industry. He's helped launch many industry-firsts including HAMR technology, 10K-rpm and 15K-rpm hard drives, drives designed specifically for video and for gaming, Serial ATA drives, fluid dynamic HDD motors, 60TB SSDs, and MACH.2 multi-actuator technology.