The long-term goal of most carmakers right now is getting to what’s called “level 5” autonomy. That’s not Tesla’s mislabeled “full self-driving,” which clearly and strongly warns drivers that they have to maintain attention on the road. Ahem. It’s not GM’s SuperCruise or Ford’s BlueCruise, either. Level 5 autonomy would actually enable the car to drive itself, without any interference from the car’s owner. Waymo is doing this already in several cities. But the idea of all private cars getting to this stage is a dream precisely because it could potentially have so many benefits:
- Elderly people who can no longer drive comfortably, safely, or easily wouldn’t have to
- DUIs would theoretically disappear, because you wouldn’t be tempted to drink and drive
- Distracted driving would also not be a hazard; you could just quit driving and sit in the back and text or work
- Potentially, municipalities would open up ride-sharing membership for cost-pooling of private car “collectives” to greatly lower the cost of transportation.
You will note that none of these mentions traffic. That’s because unless or until every single car on American roads has level 5 technology, with all cars talking to each other (V2V) and talking to infrastructure (V2X), a mixed fleet of autonomous and non-autonomous cars doesn’t resolve traffic snarls. Even 10,000 Waymo cars operating in L.A. or D.C. will not undo the stop-and-go snarls in either city.
But what if there was a technology that you could retroactively add to your own car that could do such a thing? That’s the theory behind an experiment Nissan has just tried in San Francisco, and here’s how it went.
Nissan
- Founded
-
26 December 1933
- Founder
-
Yoshisuke Aikawa
- Headquarters
-
Nishi-ku, Yokohama
- Owned By
-
Publicly Traded
- Current CEO
-
Ivan Espinosa (as of 1 April, 2025)
The Experiment
Nissan’s system is called Cooperative Congestion Management (CCM). It used Nissan’s own ProPilot Assist technology to network several cars together. ProPilot Assist already works, like Ford’s Blue Cruise and GM’s SuperCruise, allowing hands-off, eyes-on assisted driving. In some parlance, it’s considered advanced Level 2 driving or Level 2+. Whatever you call it, it’s not curing traffic all by itself.
But using CCM, Nissan networked several cars with ProPilot Assist to “talk” to each other. They created modeling and software that one of the leaders of the project, Zvi Guter explained, was designed to study live traffic based on what the cars’ assist systems were sensing. They also gathered live data from the highway system.
“This allowed us to determine what is the optimal speed for the road, given the traffic conditions, etc.”—Zvi Guter, Senior Manager of Mobility Research, Nissan
But Nissan also wanted to address not just the optimal speed, but what speed might be best for a column of cars tethered together to travel at, to actually proactively ease congestion.
The Key Component — We Drive Badly
Human beings, Guter explained, tend to follow too closely in traffic, then brake too hard in response to a slowdown. This leads to a harmonic—a pulsing wave. We speed up, then slow down, speed up, then slow down. This is an inefficient form of transit. It’s both slower than a consistent speed, and more dangerous, because if you brake harder than the driver behind you is anticipating, you’re going to be rear-ended.
Guter said CCM’s technological solution is the opposite of this harmonic, broadcasting the ideal constant speed to all the cars in the train:
“You want to minimize acceleration deceleration, right? You want your speed to be as constant as possible.”—Zvi Guter, Senior Manager of Mobility Research, Nissan
If you could smooth out harder braking and harder acceleration, in theory, you could disturb the wave of traffic along the way, too. In essence, like dropping a stone in a pond, you create a different ripple effect than the prevailing wave.
How It Worked
While computer simulations showed CCM could cut travel times by 18 percent and drivers could save up to 42 percent on gas—which is massive—Nissan took the lab experiment to the field, and tested out CCM on a busy, crowded section of Interstate 680 in the San Francisco Bay Area, and experimented over multiple sessions and a total of 600 miles.
Using the ProPilot Assist cars tethered electronically, Guter said the cars were being fed information from the cloud, but also from a lead “probe” vehicle that went out and fished in the greater pond, roughly 30 seconds ahead of the pack. That car beams data back to the following group of tethered cars, enabling them to more gently adjust their speed, rather than creating their own stop-and-go harmonic.
The profound zinger is that Nissan found the cars in the pack spent 70 percent less time stopped in traffic. A further benefit is that because they slowed much more gradually, that meant 85 fewer hard-braking incidents, and in turn, slower braking is less likely to trigger a following car not using CCM to have to hard brake, which leads to changing the overall rhythm of traffic on a given stretch of road.
The Hard Part — The Tech Rollout
Nissan scientists have been noncommittal about simply unleashing the tech in the wild, but in theory, Guter said, the carmaker could license the technology to work with any driver’s cell phone, and to broadcast the ideal speed to every driver on the road, or to subscribers to an app in a system like CarPlay or Android Auto. He said that this kind of technology wouldn’t be that difficult to roll out, because most drivers have cell phones already. Guter thinks it would be better to have it tied into a car’s cruise control, but the other option is to have a system such as Waze enable an “ideal speed” with the app, to advise the driver of that speed and perhaps incentivize that behavior in some way.
Speed monitors around school zones already do something similar, too, so smart signage might also work to ease congestion.
Some Slack In The System
When asked if the system could still work via cell phone or beamed to an app in CarPlay, say, with a set speed displayed to your phone or an app, and what happens if that speed changes, Guter made something clear: It’s not going to change that frequently or jump around. Meaning, if you’re thinking the set speed might be 45 MPH one second, then ramp to 60 MPH the next, that’s not accurate. The reality is that the carrying capacity is already modeled and factored in, so he said there could be more like a 1 MPH change every few minutes, and the wave of cars just has to be consistent.
“The slowing happens slowly, the acceleration happens slowly. That’s really the key to the puzzle.”—Zvi Guter, Senior Manager of Mobility Research, Nissan
How We’ll Screw It Up For Everyone Else
The most interesting part of this study was that Nissan demonstrated that just a few cars on a train can change the pattern of drivers around them. If all of them are moving steadily, every car behind them can also move steadily, and in turn, all of those cars will get there faster and be engaged in less stop-and-go. Everyone saves gas; everyone gets to their destination faster. It doesn’t require autonomy. It doesn’t even require all cars to have a cell phone or app solution such as CCM.
But as anyone who’s driven with adaptive cruise control can attest, those systems operate with a minimum safe stopping distance. So if you’re following safely, there’s likely at least one car space of room between your front bumper and the next car’s rear bumper. And Guter made the point that drivers who want to change lanes, pass, aggressively merge and so on could reintroduce the stop-and-go harmonic.
But he believes that if CCM was implemented, its benefits would be obvious: “You were late to work yesterday, and today you’re not.”
Not Just Highways
While Guter cautioned that Nissan hasn’t studied this application, he noted that secondary road light sequencing could also benefit from this kind of cloud-based traffic analysis and traffic flow control. A lot of roadways in the U.S. still have “dumb” stoplights that only operate on timing, and that’s in part due to the cost of traffic analysis. But cars beaming data about traffic flow to the cloud and back to stoplights would mean that stoplights could be “live” sequenced, so that they’re responding in real time to traffic.
A better version of this would also be “live” speed limit signs, because just like with using an app or ProPilot-based CCM, you’d be told the ideal speed to hit all the green lights, again easing traffic.
TopSpeed’s Take
One obvious benefit of CCM would be that it wouldn’t be vehicle-dependent. We’ve already seen brands like Volvo and Mercedes-Benz talk about car-to-car traffic and accident information, but unless you’re a driver in Sweden or Germany, with very deep brand penetration, that data doesn’t do you much good. Nissan’s solution theoretically could work regardless of what car you drive, and even work with an older car that doesn’t have an onscreen version of Apple CarPlay or Android Auto, but you could still access the data bounced to your cell phone, ideally dash-mounted, so you’d see it in an app like Waze.
Lo-fi solutions are going to be critical to cure traffic, because the average American car is nearly a decade old. With rising car prices, that won’t change. So Nissan’s tech might be just the solution we can all use.