A critical factor in determining the whether an autonomous car can perform its intended function is the setting it’s operating in. Level 5 autonomy, the highest level, is defined as a car that can operate without human input in any environment. Level 4 encompasses the same capability while confined to a specific area such as a certain highway or city. On the technology side, we are well on our way to level 4 autonomy. Ford for one has claimed that it has a level 4 car that will operate in a geo-fenced area that is heavily 3D mapped as soon as 2021,[i] and we’ve been seeing autonomous concept cars at auto shows for years. The question we all want the answer to, though, is when autonomous cars will actually be on the roads, and the general consensus is that there is no general consensus. Maybe the answer lies not in when the cars will be ready for road use, but when the roads will be ready for autonomous cars.
In a bit of a media blunder at the 2016 auto show, a Volvo XC90 carrying the brand’s CEO as well as the Mayor of LA was attempting to semi-drive itself using its Pilot Assist function. It was unable to read the faded lines of the road though, which lead to a jittery stop and start sequence as the car repeatedly refused to drive itself, prompting the Volvo CEO to exclaim, “It can’t find the lane markings! You need to paint the bloody roads here!” while the world watched.[ii] According to the Department of Transportation, 65% of American roads are in poor condition. Potholes, misleading signs, poorly painted road lines, and variations in road signage across the country work together to make it much more difficult for a robotic car to know where it’s going. Some roads, like the one pictured here, don’t even have painted lanes.
Furthermore, has the automotive industry considered how it will deal with overzealous Tufts Jumbos who like to show their school spirit with spray paint? Consider the picture above. This problem is also forcing the companies building autonomous cars to develop more advanced cameras and sensors to deal with the confusing and misleading information they pick up as they drive, ultimately pushing back the date when we can see these cars in public use, not to mention the increases cost to the consumer as a result of this extra R&D. Given these challenges, there is a slew of companies developing complex radar and lidar technology like Mobileye and Velodyne that allow cars to interpret their surroundings on their own. But what if there was a way to simply give the cars this information without the complicated and expensive sensory technology?
Enter the smart highway. Think of it as a way of incorporating technology directly into our infrastructure, providing cars with useful information that would otherwise come from these sensors or cameras. It seems confusing when explained without context, so it may be best to look at some examples to better grasp what a smart highway is. One proposal by a Dutch design lab called Studio Roosegaarde incorporates special paint that signals the conditions on the road, such as temperature sensitive paint that only appears when it is below freezing, alerting drivers or the cars themselves that the road could be icy.[iii] They also proposed dynamic road lines that change to most efficiently direct traffic depending on its flow. Images of these propositions can be seen below.
These concepts are less groundbreaking even than some of the technology already in the works though. The Korea Advanced Institute of Science and Technology has built roads with magnets beneath the surface that can charge electric cars as they drive though induction, and buses with this technology are on Korean roads today. What about here in the US? At the forefront of smart road technology is a 35-mile stretch of interstate in Ohio named the 33 Corridor after the road that contains it, US 33. This multi-group effort that includes The Ohio State University, the Transportation Research Center, and Wind River (an Intel subsidiary) involves a series of fiber optic sensors placed under the road that relays data to aid with vehicle-to-vehicle, vehicle-to-infrastructure, connected cockpit units, and smart mapping technologies.[iv] It’s first two stages of construction are already complete and it is due to be fully finished by October.[v] This road has broad implications for how self driving cars are tested. Autonomous car manufacturers now have a physical location where they can test their products’ ability to interact with smart roads well before the technology finds its way into infrastructure through traditional government funding, which tends to be slower and could in turn delay that magic date even further.
Ohio isn’t the only one jumping on the smart highway bandwagon though. In 2015 the Department of Transportation launched its Smart City Challenge where mid-sized cities from across the country would submit ideas for, as the DOT puts it, “an integrated, first-of-its-kind smart transportation system that would use data, applications, and technology to help people and goods move more quickly, cheaply, and efficiently.”[vi] The winner would receive a $40 million investment from the DOT to put towards implementing the proposed smart infrastructure. The very fact that this contest happened is a strong indicator that the federal government is getting ready for the autonomous car storm and encouraging smaller leadership bodies to get involved too. The 78 applications that the contest received show that the encouragement worked. Columbus, Ohio was the winner, whose “comprehensive, integrated plan addressing challenges in residential, commercial, freight, and downtown districts using a number of new technologies, including connected infrastructure, electric vehicle charging infrastructure, an integrated data platform, autonomous vehicles, and more” is graphically detailed below.[vii]
This plan shows that the implications of autonomous driving and smart roads go far beyond surface benefits like speed and convenience. The technology on the horizon in this field has potential to make massive changes in areas like social justice and mobility and environmental sustainability as well. The other cities that competed also delivered good plans that show just how critical smart roads will be for the surge of technological modernization that is likely to occur in American cities in the coming years. Here are some of the other propositions:
Similarly to the DOT, the National League of Cities released a policy preparation guide for autonomous vehicles this past April which highlighted the need for cities to encourage and participate in smart infrastructure investment.[viii] It’s clear then that within certain circles of the urban planning world, smart infrastructure is a buzzing topic, and it ought to be in any discussion of autonomous vehicles as well.
Many of the decisive obstacles in the way of level 4 or 5 autonomous vehicles being on the road today can be removed if we put as much into the roads these cars will drive on as the cars themselves. The need for complex sensors and data processing in each individual car will be greatly diminished if the roads can bear some of these informational burdens. While the construction needed to implement this technology may make traffic hellish for a few years, the long-term benefit of investment in smart infrastructure will pay it back many times over.