October 2018

The rapidly approaching deployment of commercially available CAVs has led city planners to begin grappling with the ways in which this new technology is expected to shape our built environment.  A 2017 report from MIT’s Urban Economics Lab and Center for Real Estate, financed by Capital One, explores potential real estate changes driven by CAVs. The report describes two theories of what the effect will be. First, CAVs could reinforce demand for central city living by relieving congestion and need for parking, making cities more livable. Alternatively, they could lead to a new wave of suburbanization by increasing the distances people are willing to travel.

As much as CAVs will shape the future of cities though, design choices made by city planners today will also impact the ways in which CAVs are utilized. Cities that are designed primarily for drivers, with limited walkability and few public transit options, are likely to experience a rehash of all the problems with 20th century suburban sprawl: congestion, increasing infrastructure needs on the urban fringe, and a reduced tax base within city limits, to name a few. There are, however, affirmative steps that cities can take to disincentivize sprawl in favor of growth in the urban core. Two of these policy options, which I will discuss below, are smart pricing of vehicle travel and increased walkability of city centers.

Many cities have already taken steps to make solo trips in cars less attractive. Whether these policies take the form of increasing options for light rail and other public transportation, designating carpool lanes, or varying parking costs depending on the time of day, many of them may not be significantly altered by the arrival of CAVs. One change that could be facilitated by CAVs is the possibility for more fine-grained trip pricing. A city that is committed to reducing congestion could vary ride pricing for people who carpool, or for trips made outside of the heaviest use periods. Those hoping to incentivize public transit could provide reduced fares for “last mile” trips to and from light rail or bus stations.

The prevalence of CAVs will also provide cities an opportunity to rethink the design of their urban landscapes. Most American cities are dominated by parking, with 30% of the space in many downtown areas being taken up by parking spaces. This is unsurprising in light of the fact that the typical car is parked around 95% of the time. The rise of CAVs will provide cities with an opportunity to adapt much of this space to more productive use through business development, building downtown housing, and expanding green space. A key challenge here for cities will be in managing the transition. A study by the Regional Planning Association for New York, New Jersey and Connecticut found that land use planning is unlikely to be “permanently altered” by CAVs until 2040 and beyond. In the intervening years, cities can begin to take steps to plan for adaptive reuse of space. This includes such design choices as building parking garages with features that allow them to be easily converted into housing and considering zoning changes that will facilitate a more livable, walkable urban core.

CAVs have the potential to contribute to the continued revitalization of city centers through the creation of more resident-friendly downtowns, or to kickstart an accelerated urban sprawl. Smart, data-driven trip pricing and infrastructure designed to smooth the transitioning needs of cities can help guide the use of CAVs in ways that facilitate compact growth and walkable communities.

Two recent news stories build interestingly on my recent blog post about CAVs and privacy. The first, from Forbes, detailing law enforcement use of “reverse location” orders, where by investigators can obtain from Google information on all Google users in a given location at a given time. This would allow, for example, police to obtain data on every Google account user within a mile of a gas station when it was robbed. Similar orders have been used to obtain data from Facebook and Snapchat.

Look forward a few years and it’s not hard to imagine similar orders being sent to the operators of CAVs, to obtain the data of untold numbers of users at the time of a crime. The problem here is that such orders can cast far too wide a net and allow law enforcement access to the data of people completely uninvolved with the case being investigated. In one of the cases highlighted by Forbes, the area from which investigators requested data included not only the store that was robbed, but also nearby homes. The same situation could occur with CAVs, pulling in data from passengers completely unrelated to a crime scene who happen to have been driving nearby.

The other story comes from The Verge, which covers data mining done by GM in Los Angeles and Chicago in 2017.  From the article:

GM captured minuted details such as station selection, volume level, and ZIP codes of vehicle owners, and then used the car’s built-in Wi-Fi signal to upload the data to its servers. The goal was to determine the relationship between what drivers listen to and what they buy and then turn around and sell the data to advertisers and radio operators. And it got really specific: GM tracked a driver listening to country music who stopped at a Tim Horton’s restaurant. (No data on that donut order, though.)

That’s an awful lot of information on a person’s daily habits. While many people have become accustomed (or perhaps numb) to the collection of their data online, one wonders how many have given thought to the data collected by their vehicle. The article also points out scale of the data collected by connected cars and what it could be worth on the market:

According to research firm McKinsey, connected cars create up to 600GB of data per day — the equivalent of more than 100 hours of HD video every 60 minutes — and self-driving cars are expected to generate more than 150 times that amount. The value of this data is expected to reach more than $1.5 trillion by the year 2030, McKinsey says.

Obviously, creators and operators of CAVs are going to want to tap into the market for data. But given the push for privacy legislation I highlighted in my last post, they may soon have to contend with limits on just what they can collect.

~ P.S. I can’t resist adding a brief note on some research from my undergraduate alma mater, the University of Illinois. It seems some researchers there are taking inspiration from the eyes of mantis shrimp to improve the capability of CAV cameras.

 

By the end of this year, Alphabet subsidiary Waymo plans to launch one of the nation’s first commercial driverless taxi services in Phoenix, Arizona. As preparations move forward, there has been increasing attention focused on Arizona’s regulatory scheme regarding connected and automated vehicles (CAVs), and the ongoing debate over whether and how their deployment should be more tightly controlled.

In 2015, Arizona Governor Doug Ducey issued an executive order directing state agencies to “undertake any steps necessary to support the testing and operation of self-driving vehicles” on public roads in the state. The order helped facilitate the Phoenix metro area’s development as a key testing ground for CAV technology and laid the groundwork for Waymo’s pioneering move to roll out its driverless service commercially in the state. It has also been the target of criticism for not focusing enough on auto safety, particularly in the aftermath of a deadly crash involving an Uber-operated CAV in March.

As the technology advances and the date of Waymo’s commercial rollout approaches, Governor Ducey has issued a new executive order laying out a few more requirements that CAVs must comply with in order to operate on Arizona’s streets. While the new order is still designed to facilitate the proliferation of CAVs, it includes new requirements that CAV owners affirm that the vehicles meet all relevant federal standards, and that they are capable of reaching a “minimal risk condition” if the autonomous system fails.

Along with these basic safety precautions, the order also directs the Arizona Departments of Public Safety and Transportation to issue a protocol for law enforcement interaction with CAVs. This protocol is a public document intended both to guide officers in interactions with CAVs and to facilitate owners in designing their cars to handle those interactions. The protocol, issued by the state Department of Transportation in May, requires CAV operators to file an interaction protocol with the Department explaining how the vehicle will operate during emergencies and in interactions with law enforcement. As CAVs proliferate, a uniform standard for police interactions across the industry may become necessary for purposes of administrative efficiency. If and when that occurs, the initial standard set by Waymo in Arizona is likely to bear an outsized influence on the nationwide industry.

Critics have called the new executive order’s modest increase in safety requirements too little for such an unknown and potentially dangerous technology. Even among critics however, there is no agreement as to how exactly CAVs should be regulated. Many have argued for, at minimum, more transparency from the CAV companies regarding their own safety and testing procedures. On the other hand, advocates of Arizona’s relaxed regulatory strategy suggest that public unease with CAVs, along with the national news coverage of each accident, will be enough to push companies to adopt their own stringent testing and safety procedures.

This more hands-off regulatory approach will get its first close-up over the next few months in Arizona. The results are likely to shape the speed and direction of growth in the industry for years to come.

 

For many people, syncing their phone to their car is a convenience – allowing them to make hands-free calls or connect to media on their phone through the car’s infotainment system. But doing so can leave a lot of data on the car’s hardware, even after a user believes they have deleted such data. That was the case in a recent ATF investigation into narcotics and firearms trafficking, where federal law enforcement agents were issued a warrant to search a car’s computer for passwords, voice profiles, contacts, call logs, and GPS locations, all of which they believed had been left on the car’s on-board memory. While it’s uncertain just what was recovered, an executed search warrant found by Forbes claims the information extraction was successful.

While this case doesn’t necessarily raise the same issues of government access to data found in the Supreme Court’s recent Carpenter decision, it does illustrate the growing amount of personal data available to outside actors via the computer systems within our vehicles. And while the 4th Amendment can (usually) shield individuals from overreach by government, personal data represents a potential target for malicious actors, as shown by the recent data breach at Facebook which exposed the data of 30 million users. As cars become yet another part of the greater “internet of things,” (IoT) automakers have to confront issues of data protection and privacy. Security researchers have already began to prod vehicle systems for weaknesses – one group was able to breach the computer of a Mazda in 10 seconds.

There has of late been a great deal of talk, and some action, in Washington, Brussels, and Sacramento, towards mandating greater privacy and security standards. Earlier this month, the Senate Commerce Committee held a hearing on Data Privacy in the wake of the European Union’s General Data Protection Regulation, which took effect in May, and California’s Consumer Privacy Act, which was passed in June. Last month, California also passed a bill that sets cybersecurity standards for IoT devices – and there are similar bills that have been introduced in the House and Senate. While it remains to be seen if either of those bills gain traction, it is clear that there is an interest in more significant privacy legislation at the state and federal level, an interest that has to be considered by automakers and other CAV developers as CAVs move closer and closer to wide-scale deployment.

Last week’s release of the Intergovernmental Panel on Climate Change (IPCC) special report highlights the “rapid and far-reaching” societal transformations required in order to limit warming to 1.5, or even 2 degrees Celsius. A new study by researchers at the University of Michigan, published in the journal Environmental Science & Technology, highlights the role of connected and automated vehicles (CAVs) in ushering in a low-emissions future. This research sheds new light on a largely understudied aspect of the coming CAV revolution. In my first post for the Journal of Law and Mobility, I will summarize that study and provide key takeaways for policymakers.

The Michigan study identifies several factors that will cause CAVs to emit more greenhouse gases than comparable human-driven vehicles. The weight of sensors and the computer system necessary to operate an CAV, the power consumed by the computer system, in particular the mapping function used to create high definition charts of the car’s surroundings, and the increased drag from cameras and sensors mounted on the outside of the vehicle, all operate to increase emissions. Depending on the weight of the equipment and power usage of the computer, these factors were found to increase emissions by between 2.8% and 20% relative to a comparable human driven car.

These factors are expected to be partially, if not entirely, offset by the car’s ability to create more favorable traffic patterns and identify more efficient routes. This extra efficiency is expected to more than offset any increase in weight, power usage, and drag under some scenarios, and turn a relatively large emissions increase into a more modest one under others. All things considered, the study finds that the emissions impact of CAVs will range from a 9% reduction to a 5% increase.

Many of the obstacles to reducing CAV emissions are engineering challenges: reducing the weight and power consumption of computer systems, and improving the aerodynamics of external sensors. Policymakers role in solving these challenges are likely to primarily take the form of support for university research and/or tax credits for private sector research on improving the efficiency of CAVs.

Policymakers do however have a significant role to play in improving the network effects of automated vehicles, such as reduced congestion. After the proper levels of safety, security, and reliability are obtained, a high volume of CAVs on the road increases the efficiencies that can be gained through cars communicating with each other to ease the flow of traffic. Laws that ensure high standards for data privacy and CAV safety can give consumers the confidence needed to use CAVs at a higher rate. Regulatory schemes that ease the entrance of CAV fleets into a city’s vehicular landscape can promote early adoption.

Particularly as the technology advances, CAVs have a role to play in reducing harmful greenhouse gas emissions. Widespread adoption of the technology can maximize these benefits, paving the way for large fleets of CAVs that create strong network efficiencies. As the technology advances to a point of being safe for public use, policymakers should account for these potential benefits as they consider the advent of CAVs in their cities.

Getting on the Road

Hello! My name is Ian Williams, and I am the Managing Editor of the Journal of Law and Mobility. I wanted to take a moment to introduce myself, and give a brief preview of what to expect from the Journal in the coming weeks and months.

As Managing Editor my job is to keep the trains, err… connected and automated vehicles… running on time. I work with the editorial board to review and edit our articles, and will be a frequent voice on this blog. Starting today I will also be joined on the blog by some brilliant law students who are serving as Research Editors for the Journal.

In the coming months the Journal will begin to publish articles and blog posts on a more frequent basis, with the help of our board and our students. We’ll also be holding events on campus at the University of Michigan, and will host a symposium next year (details to follow). If you’re interested in following our work, subscribe to our mailing list, and take the time to follow us on Twitter, @FuturistLaw.