Vehicle Automation versus Connectivity, and What it Might Mean for Traffic

The U.S. Department of Transportation’s depiction of connected vehicles on a controlled-access highway.

ed’s note: This week, we’re featuring a short series of articles from our board member Juan Matute on what he’s thinking about technology and transportation.

I have the opportunity to be involved in a lot of interesting research as the Associate Director of the UCLA Institute of Transportation Studies. Over the past year, I’ve been involved with our study of policy, behavioral, and market research to better understand opportunities and challenges for connected vehicles implementation.

The study’s formal title is NeTS:Large: Collaborative Research: Closing the loop between traffic/pollution sensing and vehicle route control using traffic lights and navigators, so we call it Green City Transportation Architecture. It’s funded by the National Science Foundation and is more exploratory than applied. Don’t expect to see the research project’s results become commercialized in the near term.

We’ve supported a team of computer scientists looking to optimize vehicle traffic flow within cities and regions through the use of connected “smart” traffic signals, a central navigation server (think Waze Plus), and a dynamic congestion charge.  Implementing such a system requires vehicle-to-vehicle and vehicle-to-infrastructure connectivity, which is the focus of this article.

Vehicles must be aware of their environment in order to respond to it. Automated but unconnected vehicles are limited to one-way line-of-sight scanning to assess their environment. Connected vehicles use data connectivity to communicate with infrastructure and other vehicles, including those outside the line-of-sight, either around curves or more than one vehicle ahead.  Connectivity enables data communication for an activity that’s largely dependent on visual communication (presence of vehicles, stop signs, lane paint, etc.), aural communication (honking), and a set of rules of the road. We’ve already seen the possibilities of data communication between vehicles (or their occupants’ smartphones) and central servers, but there are much greater possibilities from vehicle-to-vehicle and vehicle-to-infrastructure connectivity.

Many of the benefits that people associate with vehicle automation actually come from vehicle connectivity. Adaptive cruise control, where a trailing vehicle automatically speeds up or slows down to maintain separation from the vehicle in front of it, is a vehicle automation feature. However, the biggest increase in vehicle throughput only comes from adaptive cruise control with multi-car platoons, which requires vehicle connectivity

The benefits from vehicle automation accrue primarily to drivers. At higher levels of full vehicle automation, drivers can give less focus to the driving task, enabling them to focus on other tasks. With partial vehicle automation, features such as automated cruise control, lane maintenance, automatic overtaking, and parallel parking are primarily for driver convenience. The driver is expected to maintain focus on the driving task and be capable of retaking control of the vehicle at any time.

A few researchers are looking into whether or not drivers maintain focus on the driving task as they yield more control to the vehicle’s automated features. If drivers of the first partially-automated vehicles are less focused on the driving task, it’s probable that early experience with partial vehicle automation will show higher-than-expected rates of accidents. I’m told that the most difficult task for full vehicle automation is driving in a complex urban environment complete with pedestrians and cyclists – a dynamic which could bring a new dimension to active transportation advocacy in the future.

Selling drivers partial or full automation is easier than selling vehicle connectivity. New car buyers already see adaptive cruise control and automatic parallel parking as amenities that they can benefit from. It’s harder for a new car buyer to see benefits in vehicle connectivity if infrastructure and other vehicles are not yet connected. Governments need substantial funding in order to roll out connected infrastructure, especially connected stoplights that broadcast phase data so that a vehicle knows a light will soon turn green. Many of the system-wide throughput increases that come from a high saturation of connected vehicles will not be experienced in the near-term, as unconnected vehicles will remain on the streets for at least 20 years.

So what may happen if individual benefits accrue to drivers, as they buy partially-automated vehicles in the short-run, but capacity increases only come from major investments in vehicle connectivity in the long run?  That sounds like a recipe for more traffic congestion.  This result would be especially true if drivers in congested areas seek out the lane maintenance and adaptive cruise control capabilities of newer vehicles in order to delegate more of the congested driving task over to the vehicle.  Such drivers would become less bothered by congestion (they could perform other tasks in the car that require only partial attention) and possibly respond by consuming additional congested travel.

We’ll be talking about vehicle automation at UCLA’s Digital Cities: Smarter Transportation Forum in downtown Los Angeles on Thursday, March 20th.  Streetsblog Los Angeles readers can register for $129 using discount code “sbla”.

  • madelinebrozen

    As someone who rarely thinks about policy and technology on highways, and focuses on urban streets, I’m increasingly more interested in understand how the technology improvements going towards automated cars on highways can be applied to improving safety on urban streets. What are the connections here or am I grasping at straws?

  • rickrise

    Sounds like a way to cement our dependence on single-occupancy vehicles, resulting in, if nothing else, ever greater land use arrogated to driving.

    And of course computerized systems break down and lose connectivity all the time, including the smartphones touted in the article. Even announced routine maintenance can lead to unexpected system failure cascades, such ass the radar shutdown that led to the crash of a multimillion dollar automated aircraft on landing at the San Francisco airport.

    Cars themselves are the problem; even if they run on pixie dust and are steered by infallible spirits, they simply take up too much space to allow room for a healthy human society, let alone the rest of the world’s living communities.

    And, will they work if cyclists and walkers don’t themselves wear transponders? Will we eventually ban non-motorized travel de jure, as we have done de facto by car-only infrastructure, the demand for which will increase as everyone starts to see their personal car as a “free” taxi?

  • Joe Linton

    I think that these are very different than the scope of the study that Juan is profiling in this article, but I think that there’s a lot of potential technological gizmos that help with urban streets. It’s not networked, but here are some of the techno stuff I’d like to see:
    – Airbnb type applications for car-sharing / ride-sharing
    – More sophisticated realtime bike-ped-transit trip-planner stuff (where your phone tells you that your bus connection is running late, so you need to walk fast, or modify your trip.)
    – Collision-avoidance stuff (sounds more difficult after reading Juan’s article – especially technology that helps cars avoid hitting peds and bikes)

    I mostly this article as telling me what we, who care about urban streets, are up against.

  • Kenny Easwaran

    I wonder if something like Metro’s Express Lane feature might be a relevant way to do this. At this point, we’ve got lanes on a few expressways that are legally usable only by vehicles that have connectivity features (in this case transponders). Because those lanes are valuable enough, lots of people volunteer to get the connectivity feature for their car. Once we have enough people with those connectivity features, it becomes possible to start charging them to use those lanes, and we can expand the same program to other places where a congestion charge would be useful. At this point, there are probably legal restrictions on the ability of Metro or the city to impose congestion charges on the 405 or 101 (though we sorely need it). Perhaps when enough people have these transponders, or cars start coming with communications features built in, we can start charging cars that use highly congested surface streets as well, or surface streets that ought to be prioritized for bicycle and pedestrian travel.

    I see this communications technology as a source of value for everyone who wants to get around in an urban environment. For people in cars, it can help direct them to the best ways to get where they’re going (through a combination of information about uncongested streets and pricing on congested ones), and for people not in cars it can be used to help give cars an incentive to stay away from certain areas that ought to be prioritized for people.

  • Todd Litman of the Victoria Transport Policy Institute did an excellent overview of driverless autos that bears looking at:

    John Ulloth, a local activist, once pointed out to me platooning etc. so we cram autos on the freeways means dumping huge volumes of vehicles onto surface streets. Isn’t that idea built on making our streets mini-freeways. UGH!

  • andrelot

    “Computerized systems” will break down or have off-time depending on their specs, redundancy and requirements. Smartphones and personal computers cope with the occasional shutdown or, in old days, BSOD because the lower prices derived from their construction is compensated by lessened reliability.

    On the other hand, you have “computerized systems” that are extremely reliable, like those present on drones, used to manage power systems, and to control airplanes, spacecrafts etc.

    It all boils down to the adequate resilience and reliability ratios to cost.

    Even keeping the discussion within transportation, high-speed trains in Europe and Japan are essentially controlled by “computerized systems”, and they are very reliable systems. If you take a flight, ever, rest assured “computerized systems” are doing much of the flying if your airplane is a jet and was manufactured after 1996.

  • andrelot

    Collision avoidance is one benefit. Google Cars (which pertain to another category of teched-up vehicles) can recognize obstacles such as pedestrians, bicycles and even pets above a certain size, and react accordingly.


Introducing Streetsblog’s Thursday Job Market

Looking for a job? If you work in transportation engineering, bike and pedestrian planning, transit planning, or transportation advocacy, you can find new opportunities on the Streetsblog Jobs Board. We just launched the board earlier this summer and every week we have new listings coming in. For employers, posting your job openings on Streetsblog gives […]

Study: High-Traffic Arterial Roads Reduce Quality of Life, Even Blocks Away

Seminal research by Donald Appleyard in the 1970s found the volume of traffic on a street affects quality of life for residents in profound and unexpected ways. For example, the amount of social contact people had with their neighbors was curtailed for those who lived on high traffic streets compared with those living on quieter streets. People even defined their “home […]

The Highway Era Is Over. When Will Our Institutions Catch Up?

The highway era is over. The construction of the Interstate Highway System is essentially complete. Americans will continue to log lots of miles on highways, but for the most part, the job of building them is over. We’ve already connected the places worth connecting by highways. The problem is that transportation agencies — especially state DOTs — haven’t caught up. In their training, organizational […]