When it comes to fully autonomous cars, Adrian Pearmine, national director for smart cities and connected vehicles at DKS Associates, is often asked, “When?”

But for Pearmine, who’s spent over 20 years designing and implementing technology and communications solutions for the transportation and transit industries, there’s a different question to consider: When will the industry move beyond the testing phase and begin broader deployments?

“We're already seeing autonomous vehicles driving around the Bay Area,” he says.

Pearmine thinks we may be decades away from cars that can navigate entire cities, but some tangible progress within four or five years is possible. Rather than transformation sweeping entire metropolitan areas, for instance, we could see sections of cities like Portland or Los Angeles prioritizing autonomous vehicle systems.

But no matter how smart the car, it’s only as advanced as the traffic ecosystem it navigates. Pearmine explains that current infrastructure isn’t built for autonomous vehicles to communicate and exchange data seamlessly. But with next-gen connectivity, we may see cutting-edge information become accessible to vehicles sooner than we think.

High-resolution GIS mapping and GPS are car features that track precisely where a vehicle is, and according to Pearmine, a series of camera, radar, and lidar sensors can add situational awareness in order to analyze surroundings. Those surroundings may include bikes, pedestrians, and other cars in a constantly changing and complex environment.

While these sensors make the car smarter, safer, and more efficient, Pearmine explains their limits: They’re still not capable of enabling vehicles to peer around the next corner and monitor what’s ahead—whether it’s a fire truck blocks away, pedestrians at an upcoming intersection, or icy conditions in the distance.

With 5G, however, there may be an answer. “All of these are types of scenarios that a car’s computer can process if it’s fully connected,” says Pearmine. “That’s why you need to connect it to a faster mobile network to augment all the sensors and the other autonomous features.”

With next-gen mobile networks in place, autonomous vehicles represent only one segment of smart traffic infrastructure. Consider the role smart traffic signals could play. For instance, today, a car camera can process a regular traffic light and determine next moves based on signal color. But tomorrow, perhaps an intelligent traffic signal could broadcast information to the car in real time, long before it approaches the intersection.

“The other parts of smart intersections are microwaves and video analytics to determine if there are pedestrians in the walkway,” says Pearmine. “The pedestrian doesn’t have to be connected necessarily, but they can be passively detected.” 

Further, the smart intersection can broadcast information to other cars. For instance, a connected vehicle network may send a variety of data points to surrounding vehicles: A car is about to turn right, its right turn signal is on, and it’s in the right-hand lane.

In this scenario, what’s most crucial is data that benefits other drivers, pedestrians and autonomous vehicles in the area—data that can ultimately preempt a collision and save lives.

Safety is already a major concern as even semi-autonomous vehicle tech emerges. Pearmine cites a fundamental building block of the connected vehicle industry’s safety efforts—the Department of Transportation’s basic safety message. The protocol informs safety applications and outlines data on connected vehicles, such as vehicle size, position, speed, heading, acceleration, and brake system status.

According to the National Conference of State Legislatures, 29 states have enacted legislation related to self-driving vehicles on state roads. More states are expected to follow, particularly as the Department of Transportation limits federal oversight of autonomous vehicles. 

As change unfolds and states react, what we may encounter is a blend of autonomous, semi-autonomous and human-driven vehicles. Robust networks with ubiquitous 5G coverage will only speed up the process. Soon enough, we’ll be able to tackle our to-do list in the back seat as our car shuttles us to the office.

But ultimately, for Pearmine, it all goes back to safety—the motivator behind his continued work in the field. “There’s this idea that cars are safer when humans are driving and making the decisions,” he says. “It’s just not true.”

Originally published on Forbes.com.