By: Mike Wolfe

A recent and illustrative example is The Mayo Clinic’s use of self-driving vans to shuttle samples from its drive-through COVID-19 testing sites in Florida to its Jacksonville laboratory—freeing up healthcare workers to attend to other tasks, while minimizing their infection risk.

But, in many ways, this headline application, which supports the case for autonomous vehicles, concurrently highlights the many limitations of today’s autonomous driving technology. The Jacksonville Transportation Authority, for example, had to block off a route specifically for the self-driving vans to eliminate both pedestrian and motor vehicle traffic. It is also running a mobile command center to actively oversee the service. Finally, the onboard technology in the vans—like many other approximations of autonomous vehicles—focuses primarily on the self-contained operation of each vehicle through a combination of sensors, GPS, odometry, and machine learning.

At present, autonomous driving isn’t yet autonomous.

In order to achieve that level of operation (read: Level 5), there must be an underlying network to support, connect, and integrate the incredible advances in real-world vehicle technology and tomorrow’s autonomous applications. 5G is the leading candidate to provide this vital infrastructure and bring autonomous vehicles into the realm of everyday reason.

The case for 5G

As a network for powering safe and effective autonomous driving, 5G has several key advantages in its favor: wireless delivery, high-capacity, high-reliability, low-latency, and slice-ability.

Wireless Delivery: It goes without saying that autonomous driving will rely on a wireless network. But the ability to provide a wireless network with the capabilities (reliability, capacity, etc.) traditionally associated with wired networks is among 5G’s chief advantages.

High-Throughput: 5G is capable of delivering throughput of up to 10 Gbps. Driving applications such as real-time, high-resolution map and traffic layouts and uploading vehicle sensor data to the cloud are expected to require a minimum of 150 Mbps, so ensuring a healthy ceiling of bandwidth to support mission critical applications is a key mark in favor of 5G.

High-Reliability: Wireless networks for autonomous driving will assume a role similar to that of a utility in managing the increasing reliance on high-speed wireless connectivity. 5G, as an evolution of previous standards of wireless connectivity, has the advantage of building on trusted network technology that also enables critical fallback to 4G as well as specifications that dictate seamless hand-offs from one 5G node to another.

Low-Latency: 5G’s potential for ultra-low latency (<1ms) make it an ideal candidate to address real-time connectivity needs. Although there is a natural tradeoff between maximizing latency, throughput, and reach, network slicing will allow 5G infrastructure to dial in connectivity and latency quality-of-service baselines for specific applications.

Network Slicing: By portioning out ‘slices’ of the 5G network to support particular use cases—such as enhanced mobile broadband, low latency with ultra-high-reliability, and machine-to-machine communications—operators can allocate resources and optimize network topology to address specific service level agreements. They can also ensure that the security requirements of V2X communications are met, supporting high-quality vehicle-to-vehicle communications as well as situational awareness applications like vehicle-to-network and vehicle-to-infrastructure communications.

Requirements beyond 5G

While 5G provides a firm foundation for building the future of autonomous driving, it will require enhancements from a variety of supplemental technologies and configurations to achieve the redundancy, reliability, and connectivity to satisfy a successful foray into the mainstream. Some of the most important include other communications technologies like DSRC, architectures like mobile edge computing, and supporting networks such as mmWave, small cell, and fiber.