Paving a Clear Road to Highly Automated Vehicles

Today, driving can be frustrating if you’re stuck in traffic—which tends to be a regular experience for many commuters. In the United States alone, 5.5 billion hours of productivity are lost from waiting in traffic each year.

So what is an automated vehicle? Given that the term means different things to different people, let’s start with a definition from Gartner: “An autonomous [or automated] vehicle is one that can drive itself from a starting point to a predetermined destination in ‘autopilot’ mode using various in-vehicle technologies and sensors, including adaptive cruise control, active steering (steer by wire), anti-lock braking systems (brake by wire), GPS navigation technology, lasers and radar.”

In the automated stage, or L2 per the SAE (Society of Automotive Engineers) defined levels, vehicles incorporate features like automated braking and cruise-control. We’re pretty much there today. At the next level, highly automated vehicles, or L3 and L4, add features like self-parking and collision avoidance—and we are already seeing advances in these disruptive technologies today. And at the ultimate driverless or fully automated stage, L5, being developed and tested today, the car can drive itself without any involvement from a human driver.

So how do we get to a world where fleets of vehicles are rolling down our streets and highways with absolutely no one in the driver’s seat or even anywhere inside the vehicle? This is exactly where we are headed, but how do we get there? Let’s look at an overview of the five requirements that must be met to create an end-to-end platform for highly automated and driverless vehicles:

In-Vehicle Compute Platform. To process the vast amounts of data in real time, the AV needs an in-vehicle computing platform that processes the data with the highest performance per watt.
High-Speed Communications. The AV requires a robust, high-speed, advanced cellular (5G) connection to the cloud that is ultra-fast, ultra-reliable, and ultra-secure. We need to build out 5G communications infrastructure to enable the AV to communicate continuously and react to data in real time while on the road and with servers in remote data centers.
Robust Data Centers. The AV can’t do it all on its own. Its capabilities are enabled and enriched by data centers that can support unprecedented amounts of data and memory-intensive deep learning models that develop ever-richer datasets that are used to continually retrain the vehicle.
HMI Interface. In the AV, a human-machine interface (HMI) is needed that builds trust between passenger and vehicle. This HMI interface will incorporate technologies ranging from high definition visual displays to interactions from voice and other multi-sensification capabilities that enrich the user driving experience.
Security. End-to-end integrated security—that stretches from door locks to the data center, from the in-vehicle on-board processors to the servers in cloud/data center—is an essential building block for the AV. This security protects the integrity of the automated functions of the vehicle and the data that is in the vehicle, in transit and in data centers, and it guards against threats like hackers who might attempt to take over control of a vehicle or invade your data privacy.

These are the basic technology requirements for AVs. In their more sophisticated forms, AVs have additional requirements for artificial intelligence that continue to augment human skills and capabilities, in the form of machine learning and deep learning. With machine learning, computer algorithms detect patterns and make predictions based on data. Machine learning allows the AV to take actions without being explicitly directed to perform specific functions.

Deep learning is an advanced method of machine learning that uses neural networks to comprehend more complex and unstructured data. Deep learning can accelerate processes like image recognition, natural language processing and other complex, data-driven tasks. These capabilities allow the AV to recognize objects in the road and understand human commands.

As we proceed down the road to fully automated vehicles, from technology to manufacturing, it’s going to take a broad ecosystem of partners working together—including world-class technology companies, automotive manufacturers and telecom providers—to develop the technologies, define and align on industry standards, integrate security frameworks, build out the advanced communications infrastructure, and build the fully automated vehicle.

This global ecosystem movement is already well under way. A few examples:

BMW Group, Intel, and Mobileye are working together to align the industry on an open, secure, standards-based automated vehicle platform to quickly bring automated vehicles to market. The ultimate goal is to create the safest highly automated vehicle platform that aligns the industry and accelerates time to market.[5]
General Motors and Lyft have entered a long-term strategic alliance to create an integrated network of on-demand autonomous vehicles in the U.S. Under this alliance, GM will invest $500 million in Lyft to help the company continue the rapid growth of its successful ridesharing service.[6]
Ford is committed to delivering driverless vehicles in next five years—with no steering wheel or pedals—to be used by commercial-fleet operators and others.[7]

As these automated vehicle partnerships suggest, we’re not talking about the distant future here. We’re talking about R&D efforts that are focused on putting AVs on the road within a few years. Many industry observers accept that fully automated vehicles will be on the market within five years.

Although there is a lot of work yet to be done, there are many reasons to be excited about the rise of the AV. Imagine a whole new driving experience that will be safer, more reliable, and more enjoyable. Imagine vehicles rolling down our streets and freeways with absolutely no one in the “driver’s seat”—or even anywhere inside.

This day is not that far off. Think 2021.

Bridget Karlin is the managing director in the Internet of Things Group at Intel Corporation.