While autonomous vehicles could provide a safer, more efficient transportation future, a number of speed bumps remain before this emerging technology can become a reality. To make the next great leap forward, what’s needed includes compute, network and storage technologies capable of meeting the rigorous demands of a self-driving future.
Some existing technology solutions can present challenges. Adding supercomputer-level power to a car would double the cost of each car making them too expensive to buy. And today’s technology isn’t fast enough to offer the near-instant reaction times needed to safely navigate dynamic events like weather or pedestrians. Finally today’s GPS location data is too imprecise, offering accuracy down to a few meters rather than centimeters.
Enter 5G and multi-access edge computing (MEC). This combined architecture is well suited to store, process, transmit and analyze gigabytes worth of autonomous vehicle data with ultra low latency and high reliability. They could be the key to transforming our roadways into a new kind of information superhighway.
Benefits of a data-driven future
Autonomous vehicles could vastly improve the future of road-based transportation. A study by the University of Michigan predicts autonomous vehicles will contribute to improvements in public health and safety, greater mobility, reduced energy and environmental impacts, and increased adoption of car sharing. The same study found that autonomous vehicles could reduce the number of fatal vehicle crashes by 90 percent just in the US, potentially saving approximately $190 billion per year.
A closer look at 5G and MEC
5G and MEC could help make autonomous vehicles a roadway reality because they can channel the massive amounts of data generated by vehicles and transform it into meaningful action.
This is a topic many automakers I’ve talked to have struggled with. They can certainly generate reams of data, but what happens next? The answer starts with the network and cloud computing at the edge.
A 5G network running on mmWave spectrum can deliver peak download speeds of up to 2 gigabits per second and upload speeds of more than 200 megabits per second -- exactly the kind of bandwidth needed for autonomous vehicles generating terabytes of data.
"When you pair 5G with MEC, vehicles no longer need high-powered computers alongside their engines and motors."
Jyoti Sharma, Verizon.
When you pair 5G with MEC, vehicles no longer need high-powered computers alongside their engines and motors. Instead, data transmitted via 5G can go to a MEC zone for the heavy lifting of processing data, enabling rapid, automated decision making for autonomous vehicles facing dynamic driving environments.
Making this all work is precise location data. Autonomous vehicles have to know their location within a few centimeters -- more than GPS alone can deliver. That’s where developments like Hyper Precise Location with Real Time Kinematics (RTK) can extend the capabilities of GPS, providing centimeter-level location accuracy that far surpasses the accuracy level of three to nine meters of standard GPS.
Tapping the future of autonomous vehicles
"The benefits of these connected technologies could extend the power of remote vehicle sensors like radar, cameras, and LiDAR in ways we haven’t seen before."
Jyoti Sharma, Verizon.
The benefits of these connected technologies could extend the power of remote vehicle sensors like radar, cameras, and LiDAR in ways we haven’t seen before. Suddenly, the power of the network connected to the Internet of Things devices should let autonomous vehicles see far beyond their immediate field of view.
Imagine a future where autonomous vehicles can share traffic alerts or provide updates on changing weather and automatically clear pathways for emergency vehicles. Because of the low lag, low latency network, remote driving of vehicles could also be possible, meaning an operator in one city could “drive” a vehicle in another part of the country.
Other possibilities include the ability to automatically optimize speeds to avoid hitting traffic or missing red lights, better fuel or battery savings, and of course increased safety in and out of the vehicle. Similar to the evolution of smartphones, automakers have already hinted at a future of software defined cars with software updates that let users “patch” their vehicles, adding new features and performance capabilities.
Driving the future, together
Companies like Honda, Harman, Renovo, Savari and LG Electronics are testing how 5G and MEC can enhance Vehicle-to-Everything (V2X) applications. Initial tests have included next-generation road-alerts to emergency vehicles, pedestrian crossings and red light runners.
These collaborations are exciting and hold much promise as we remain focused on building new technologies that help meet the unique network requirements automakers have for reliability, data privacy and security, latency and throughput.
"To truly make autonomous vehicles the future of transportation, it’s going to take a deeper collaboration between technology providers."
Jyoti Sharma, Verizon.
No one company can create the entire network and edge computing infrastructure required for a more robust autonomous vehicle future. The continued evolution of hyper-precise location services and the benefits of 5G network adoption could improve the way autonomous vehicles transfer data, respond in near-real time and more accurately view what's around them. To truly make autonomous vehicles the future of transportation, it’s going to take a deeper collaboration between technology providers, automakers and regulators to solve industry-wide challenges such as spectrum allocation and regulatory policies. 5G’s low latency and high bandwidth along with the compute power of MEC are a part of the solution, but to truly embrace widespread adoption of autonomous transportation, the investment from automakers and the support of regulators are additional necessary pieces to drive this future forward.
