Well, perhaps that’s a little bit overblown. But the promise of self-driving vehicle technology - given a recent boost by Google’s self-driving car - has attracted the attention of technologists, lawmakers, and even members of this very blog. Speculation abounds as to what level of autonomy future cars will have, when they’ll be on the roads, and how they’ll change the transportation landscape. What is certain is that vehicles will increasingly do more of the driving themselves, and this has the potential to radically alter the physical, social, and legal environment of driving in America.
Oregon needs to move quickly to pass legislation that will allow for a safe and controlled introduction of autonomous vehicles. We cannot wait idle for more federal guidance while other states seize the opportunity to shape the future of American transportation.
What is an autonomous vehicle, exactly?The phrase “autonomous vehicle” hides two important distinctions that are critical to describing the technology coming to cars of the future.
The first is the method of automation, the second is the level of automation. Self-driving technology comes from two lineages, sensor-based technology and vehicle communication technology, as described in a recent KPMG report.
Sensor-based technology equips vehicles with advanced remote sensing equipment that can capture, analyze, and act on the nearby road environment. This is the technology behind the Google cars, which use a $70,000 LiDAR (laser-based 3D radar) sensor as the backbone of the system. This technology was born in the DARPA Grand Challenge, a government-sponsored race for autonomous vehicles. Initially the race was through a desert, but later iterations put the cars through simulated urban environments.
Vehicle communication technologies involve the direct wireless communication between connected vehicles, or between vehicles and nearby infrastructure. These technologies have long been a common focus of research and development by the federal government. Most recently, the Vehicle Infrastructure Integration Coalition has organized major federal and state departments of transportation with industry groups to promote connected-vehicle technology. Dedicated Short Range Communication (DSRC) is the most promising standard, and likely to be the backbone of future implementation. DSRC has proven to be fast, economical, secure, and protective of privacy.
Both of these technologies offer many potential benefits, but will be exponentially more impactful as the technologies converge inside the vehicles of the (near) future. There are three main promises that accompany the flurry of discussion around self-driving cars: driving will be safer, easier, and more efficient.
Safer.Safety is expected to be dramatically improved as increasing amounts of control is allotted to automated driving function. Function-specific automation is already on the road, in the form of electronic stability control and adaptive cruise control. The USDOT estimates that up to 80% of all non-alcohol related crashes could be mitigated using connected-vehicle technology alone (KPMG report). In general automated functions can be ever-vigilant, responding faster than a person and monitoring information invisible to the driver. These safety increases will likely apply to a wider range of situations as technologies advance.
Three types of crash seem particularly well suited for automated mitigation: running off the road, crashes during severe congestion on highways, and avoiding objects that suddenly enter the road. Adaptive cruise control, automatic breaking, and lane centering together provide for a very effective automation scenario, described by the NHTSA as combined function automation. The NHTSA guidelines for vehicle automation describe levels 0 through 5, ranging full manual control to full self-driving automation capable of functioning without any driver input.
Easier.Level 3 and Level 4 automation, which allow for the car to self-drive under some or all conditions, offer a revolutionary potential to make driving easier. When drivers are free from active control of the vehicle, they have the opportunity for productivity and recreation. This is likely a part of Google’s end game with their self-driving technology, so that you can sit back and enjoy the Internet (brought to you by Google) during your drive. The daily commute has the potential to be much less stressful as well, which brings with it personal health benefits. With Level 4 automation, a car could drive itself to you, bringing a dramatic new flexibility to personal vehicle travel.
More efficient.Given control over highway operation, a critical mass of autonomous vehicles could achieve dramatic efficiencies, reducing the negative externalities of a given distance of travel (KPMG report). Platooning of cars, with many cars self-driving at uniform speeds and tight spacing, can dramatically increase highway capacity without any new construction. Additionally, reduced air drag in a platoon (think drafting bicyclists) can increase fuel efficiency by 20%. Vehicles with awareness of congestion levels can compensate at a system-wide level, reducing bottle-necks and improving the reliability of travel times.
Disruptive.High-end consumer cars focused on taking the driver out of much of the daily commute are likely to be the first self-driving vehicles on the market. But the potential for market disruption extends into commercial, industrial, and public sector vehicles and services.
Industries that require large amounts of constant, predictable driving have an opportunity to benefit greatly from self-driving vehicles. Imagine Google’s requirements to provide their Street View feature - a huge fleet of vehicles constantly traveling pre-determined routes. Autonomous driving could cut labor costs, increase productivity, and improve reliability. Similar opportunities exist in the freight industry, as well as for public transit providers. Carshare providers such as Car2go also have an opportunity to greatly improve their service with self-driving technology, transforming it from a car rental to a virtual towncar service.
A Dystopian Future?A constant refrain in the world of planning is to watch out for unintended consequences. This missive is especially important with a dramatic, disruptive technology. Increasing the utility of personal vehicle travel time may result in a new disregard for distance traveled, increasing the threat of sprawl. The mental and physical freedom provided could negate one of the few advantages transit currently has over driving. As a new state and federal transportation focus, autonomous cars and supporting infrastructure could disrupt the gains made in multi-modal transportation planning. And with high-profile cyber-security and electronic-privacy concerns in the news, self-driving cars could be the next frontier of crime and abuse. Only time and experience will test the truth of these concerns, but active and aware participation by governments can help to mitigate these risks.
Take the wheel.Futurism aside, autonomous vehicles need to be addressed immediately. The technology is on the road now, and advancing quickly. Yet many fundamental questions remain unanswered. When self-driving vehicles experience errors, who is liable? The technology company, the car company, the driver (if there is one), or the state infrastructure provider? When and where will self-driving cars be allowed to operate without drivers? Will there be designated lanes and routes, or will they be mixed with the crowds of human drivers?
For Oregon, it is better to embrace, encourage, and guide the development of autonomous vehicles on our roads. The state should pass legislation that accommodates and regulates limited testing, registration, and operation of autonomous vehicles at designated locations in Oregon. The legislation should also put in place a program to begin planning for the future, including community outreach.
This bill should help Oregon lead, fostering research and development in the Silicon Forest. Intel has shown a desire and potential to become a major player in these technologies, and Oregon should ensure that they can invest locally. Freight along the I–5 is a prime target for autonomous vehicle technologies. Addressing climate change, autonomous vehicle registration offers a unique vector to phase in low-carbon fuel requirements. Aware that technology, federal guidance, and public opinion are quickly evolving, the plan should incorporate frequent self-assessment and maintain flexibility.
This is an opportunity to ensure that the transportation policy innovation that Oregon is known for will be integral to the national development of autonomous vehicle policy. Pedestrian safety, bicyclist integration, social equity, and environmental concerns all deserve a place in this important policy discussion.