January/February 2016
Driving the Future
BY EVA KAPLAN-LEISERSON
These days, it’s hard to read the news without spotting an article about autonomous vehicles. The topic has captured the imagination of the media and public, not only for the technology’s potential but also due to the myriad issues still needing resolution.These include questions of safety, ethics, infrastructure, and regulation.
Just as the car revolutionized many aspects of American life beyond the way people traveled with their horses and buggies, so too do autonomous vehicles offer the potential for widespread societal change. But there is much to be determined as that shift occurs.
Numerous organizations, groups, and individuals are adding their input to shape the new future. NSPE, as the recognized voice and advocate of licensed professional engineers, has an important role to play.
The More Things Change…
Since NSPE’s founding in 1934, the Society has facilitated licensed professional engineers’ protection of the public health, safety, and welfare through the ethical and competent practice of engineering. Or, as the Society stated recently as a “Grand Challenge,” NSPE aims to engage the public in “recognizing the importance of licensed engineers in advancing and protecting the public interest.”
But as technology speeds ahead, the Society is looking beyond the built environment. To this end, the organization has begun examining the ways emerging technologies affect the public. As technologies develop, NSPE wants to emphasize the need for licensed engineers and a focus on the ethical implications.
As NSPE President Tim Austin, P.E., F.NSPE, puts it, because the Society is the only organization that represents engineers across multiple disciplines, it can serve as a voice of reason and represent the public’s concerns. “Engineers and society are looking for NSPE to provide that leadership.”
Narrowing the Focus
Before licensure existed, NSPE Executive Director Mark Golden explains, people didn’t set out to build dangerous buildings, bridges, or water systems. But the need for such basic infrastructure and the pressure to meet this demand occurred more rapidly than thoughts about the public safety implications.
Similarly, the Society believes that autonomous vehicles—and related intelligent road systems—are areas of innovation that require attention now, while the new technology is still emergent. And that attention should come from people with not only appropriate technical expertise, but also the ethical and safety sensitivities of licensed professional engineers.
According to Dan Wittliff, P.E., F.NSPE, chair of NSPE’s Committee on Policy and Advocacy, the NSPE Board first identified a concern over the effects of artificial intelligence on the as-built infrastructure. However, that issue needed to be narrowed down. The board directed the committee to consider the emerging AI issues that would most directly affect the PE, namely autonomous vehicles.
Wittliff, an NSPE past president, says autonomous vehicles were a key concern because of the ways they interface with the built environment, such as roads and bridges. “Transportation infrastructure is not designed to consider artificial intelligence and autonomous vehicles,” he explains.
The questions NSPE poses: What must society do to ensure that the same governing values of safety and reliability that went into building the physical transportation infrastructure are applied to the emerging software-based transportation infrastructure, such as smart highways and autonomous vehicles? And how can NSPE be the organizing force behind getting it done?
More Questions Than Answers
The answers on how to protect the public in these areas may not be obvious—and so many questions remain.
While some experts say that the public use of fully autonomous vehicles is still quite some time off, some cars already come equipped with driver assistance technologies—for instance, automatic lane keeping (alerts the driver when a car veers out of the lane and may make steering corrections) and adaptive cruise control (maintains a safe following distance behind the car in front).
A number of still-evolving issues need expert consideration.
A sampling:
Safety. Improving road safety is one of the benefits most touted by autonomous vehicle proponents. More than 30,000 people die each year due to driver error. Supporters say that technology advances can drastically reduce the number of road fatalities.
However, others are concerned that, with partially autonomous vehicles, driver attention may be delayed when needing to take vehicle operation back over. (The National Highway Traffic Safety Administration and SAE International, formerly the Society of Automotive Engineers, have both released taxonomies for levels of automation ranging from none to full self-driving automation, with levels of partial automation in-between.)
Google wants to jump right to full automation. In Silicon Valley, they’ve been testing vehicles without steering wheels and gas pedals.
But even fully automated vehicles would need to interact with traditional cars and drivers during a transition period. What are the safety implications, especially considering accidents with Google’s self-driving cars have been blamed on other vehicles’ human drivers? Some believe safety may worsen during the transition process. Others, like Tesla’s Elon Musk, believe human driving may one day be outlawed.
According to Bryant Walker Smith, an assistant professor in University of South Carolina’s schools of law and engineering who specializes in risk, technology, and mobility, there’s not yet a consensus for what the goal posts should be. In other words, “how safe is safe enough,” given the unacceptable status quo of roadway deaths.
It’s critical that the public understand both the potential opportunities and risks, he says. “Developers have a big role to play in appropriately managing public expectations,” and in making the safety case.
That should include what safety means to them; how they have attempted to design in safety; the measures used to assure safety; and the mechanisms to monitor, evaluate, and address safety once products hit the road, he adds. These factors will be important for the public trust.
Liability and Ethics. In the current stage of development, automated vehicles have many grey areas. For instance, will manufacturers or drivers now be liable for accidents? How will those decisions be made?
In addition, technologies that shift the onus for safety onto product developers raise ethical dilemmas, similar to the ones in the classic thought experiment known as the Trolley Problem. For example, how will an autonomous vehicle “decide” in cases when the only options lead to bad outcomes, such as a crash that would harm the driver or a pedestrian?
Still, some believe this issue has been overhyped. Smith says he is asked this question more than any other, but he believes these types of choices are rare—and in most cases can be avoided or mitigated by careful driving.
He believes the real, fundamental ethical question is, “What is the proper balance between caution and urgency in bringing these systems to market,” given the number of people who die or are injured in car crashes every year? How aggressively do we move forward on technologies that will be imperfect but could save lives?
Somewhere in the middle is the right balance, he notes.
Infrastructure. Autonomous vehicles could change the whole transportation ecosystem, but it remains to be seen exactly how. There are many questions. How will roads and highways evolve? What elements will no longer be needed (perhaps traffic signals and signage) or could change (lighting, specialized lanes, lane widths)? What new elements might be needed (a national wireless communication system)?
And who will pay for these changes, when the nation struggles to fund even basic infrastructure needs?
Cities and states must attempt to plan ahead without knowing exactly how the technologies they’re planning for will operate.
Although a number of car manufacturers say they’ll be ready to release products by 2020, only 6% of the largest US cities include language about self-driving vehicles in their long-range transportation plans, according to the National League of Cities.
“I really feel for anyone” trying to conduct transportation or infrastructure planning, says Smith. It’s important to consider the range of potential future scenarios and “show your work,” he explains. Recognize that we don’t have all the answers, but at least start asking the questions.
Regulation. Technology is advancing faster than standards and regulations can catch up, and it’s unclear what those standards and regulations should look like.
Regulation is a difficult endeavor because autonomous vehicles involve both vehicle design (regulated by the federal government) and operation (regulated by the states). In addition, a catch-22 exists, explains a May Wired article: It’s difficult to judge safety without testing, but we don’t want to test without knowing that technologies are safe.
A number of states have passed laws related to autonomous vehicles, but the approach is patchwork.
The National Highway Traffic Safety Administration has been studying the issue and related technologies. In 2013, it advised states to limit the authorization of self-driving vehicles to testing and not yet allow general use.
But according to a Department of Transportation spokesperson, "breath-taking progress" in technologies has been made in the last several years. Transportation Secretary Anthony Foxx has directed NHTSA to update its guidance both to reflect current technology and, the DOT spokesperson explains, “his sense of urgency to bring innovation to our roads that will make them safer.”
Walker points to the need for an “ongoing iterative dialogue between the law and technology, between developers and regulators of these systems.” Flexibility at federal, state, and local levels will be important to innovation, he believes.
NSPE Moving Ahead
NSPE’s Committee on Policy and Advocacy (COPA) has been examining autonomous vehicles as one of a number of key policy issues, taking it through three stages: threat assessment, policy development, and advocacy.
The threat assessment working group has now finished developing its draft that outlines concerns for the public health, safety, and welfare and the profession.
It will be reviewed by the NSPE Board, along with the draft policy statement that, according to COPA Chair Wittliff, examines “the appropriate role for the licensed professional engineer in the development of autonomous vehicles in today’s as-built infrastructure.”
In addition, he explains, the policy advocacy working group will create a one-page summary that describes the problem, the dangers of inaction, and recommended actions. After approval by NSPE leadership, the summary will be given to decision-makers.
The Society has already begun its advocacy. On December 1, NSPE and the California Society of Professional Engineers submitted a joint letter to the state’s Department of Motor Vehicles, which has been charged with creating the first state-level regulations for fully autonomous vehicles.
Partnering with CSPE is important because the issue has national implications, says NSPE President Austin.
The letter urged the CDMV to ensure that its policies “recognize and incorporate the key role of licensed professional engineers in the development of autonomous vehicles and their associated systems.”
Since PEs already play a key role in protecting the health, safety, and welfare of the public using the nation’s infrastructure, it reads, “the design, development, and deployment of autonomous vehicles and their support systems should not be treated any differently and should require the involvement of professional engineers to give the highest level of assurance for the protection of the public health, safety, and welfare.”
But the work won’t stop at just one letter, Wittliff says, describing the need to stay on point and then repeatedly sell that point. “You have to convince people.”
In this work, NSPE aims to collaborate with other organizations with shared interests and has already started informal conversations with other engineering, transportation, and intelligent technology groups.
Fast Future
In emerging technology areas, NSPE doesn’t have all the answers, says NSPE Executive Director Mark Golden. But the Society can serve as “that character from childhood cartoons, the image of your conscience, that little guy standing on your shoulder who is going to say, ‘Wait a minute, I understand you’re excited and want to just jump in and start doing this, but have you thought about…?’”
It’s not helpful to oversimplify the debate by dividing into two opposing camps: the Luddites versus the Visionaries, he explains. What NSPE can say is, “Innovation is creating these challenges. Here’s where we think it’s going, here’s where we think the PE can help.”
In 10 years, Golden hopes the Society will have been a positive force in supporting innovation and new technology—but in a way that includes an awareness of the larger impact of these developments.
“If NSPE is successful,” he says, “the Society will have had such a pervasive effect on attitudes towards how this technology is being pursued, and public safety and welfare will have become so integral an aspect of its management and deployment, that we won’t need that voice of our conscience to remind us to be cautious anymore. It will be a given.”
As Austin states, “In this time of exponential change and uncertainty, licensed professional engineers must stand taller and stronger than ever before to protect public health, safety, and welfare. We must be at the forefront of these debates to raise the public’s awareness and appreciation of who we are and the principles upon which we stand. Only then can the full value of the PE license be realized. This is but the first effort to meet NSPE’s Grand Challenge.”
Share your thoughts on the PE’s role in the development of autonomous vehicles at pemagazine@nspe.org.
Look for further coverage of NSPE’s work regarding autonomous vehicles and its Grand Challenge in PE magazine and the Society’s website, e-newsletters, and social media channels.