Buckle Up

Take the Wheel

Assisted-driving technology isn’t so new. Cruise control was invented in 1948, and the Toyota Prius has been able to park itself since 2003. Most new cars today have sensors and cam-eras that will check your blind spots, keep you in your lane and warn you when you’re about to back into that shopping cart. These innovations, combined with increasingly sophisticated GPS systems and navigation apps like Waze, have helped drivers gradually become comfortable with Levels 1 and 2 of the autonomous driving scale.

“If you look at what the automakers in Detroit were doing, it was a slow move toward partial autonomy,” says computer science Prof. Holly Yanco, director of the university’s New England Robotics Validation and Experimentation Center. “But then Google came along and broke the entire model by removing the steering wheel and pedals from the car. And then Uber broke it worse.”

She’s referring to the fact that those companies (along with others such as nuTonomy, which has partnered with Lyft to pilot self-driving cars on the streets of Boston) are skipping the baby steps and jumping right into fully autonomous Level 5 vehicles.

It’s a bold strategy, says Manning School of Business Assoc. Prof. Berk Talay, who researches auto industry innovation. “It’s really too early for that technology to take over a significant portion of the market,” he says. “It will be a huge challenge to convince people to ride in autonomous vehicles where there’s nothing you can do if something goes wrong. It’s not comfortable.”

Indeed, a recent Brookings Institution survey found that 61 percent of Americans were not inclined to ride in self-driving cars, while 69 percent of those surveyed by the Advocates for Highway and Auto Safety said they were concerned about sharing the road with autonomous vehicles. Those results were likely impacted by headlines about a self-driving Uber that hit and killed a pedestrian in Arizona and two fatal crashes involving Tesla automobiles in autopilot mode.

But more than 35,000 people die on American roads each year; that’s almost 100 per day. So aren’t a few fatal accidents involving self-driving cars, while tragic, still an improvement? Isn’t it better to take the steering wheel away from humans and eliminate deadly threats like speeding or driving while distracted, drunk or drowsy?

Nicholas Evans is an assistant professor of philosophy who studies the ethical dilemmas posed by emerging technologies such as self-driving vehicles. He’s heard people like Tesla founder Elon Musk make that argument, but Evans sees it a different way.

“There are two reasons Americans die so often in car crashes. One is that we have terrible road laws, and two is that Americans drive a ton,” says Evans, who points out that 35,000 road deaths translate to 1.18 per 100 million miles driven. Autonomous vehicles, meanwhile, have only logged about a collective 20 million miles driven. In that context, three fatalities is actually worse.

Evans is working with philosophy lecturer Heidi Furey and civil engineering Asst. Prof. Yuanchang Xie to research the ethical dilemmas created by self-driving cars. For instance: Should your car be programmed to drive you into a ditch (likely causing you harm) to avoid injuring others? The team won a three-year, $556,650 grant from the National Science Foundation to come up with ethical answers to these types of questions, translate them into decision-making algorithms and then use computer modeling to see how the algorithms play out in various high-risk scenarios. 

“It’s an open question as to how we should program autonomous vehicles,” Evans says. “In the past, if you were in a crash, you acted on instinct and weren’t blamed for something that you had no control over. But with autono-mous vehicles, we do have control now. We have to program in what happens when there’s no time to think, and that’s a thorny decision.”

UML's Impact on the Future of Cars and Driving

Road ScholarS

If self-driving cars can eliminate bad drivers and traffic-snarling accidents, does that mean roads of the future will buzz along smoothly, free of congestion? Not so fast, according to Xie, who thinks autonomous vehicles could actually make our traffic worse. “People may not care if they’re stuck in traffic for an hour if it’s like sitting on their couch and watching TV,” he says. “It’s really difficult to predict when you’re dealing with humans.”

Assuming self-driving cars would be programmed to obey speed limits, however, that itself could help to ease traffic congestion. According to Evans, when London lowered the speed limit on one of its main highways from 60 mph to 40 mph, people’s commuting time was reduced by a third. Instead of cars constantly accelerating and braking (which is also bad for fuel economy), they cruised along at a steady speed.

As part of the research on the ethics of self-driving cars, Xie is using traffic simulation software called PTV VISSIM (“Because the NSF wouldn’t buy us a Tesla,” Evans says) to see how autonomous vehicles can strike a balance between safety and efficiency, particularly on freeway ramps and at intersections.

“If autonomous vehicles are going to be extremely cautious, they’re going to be very slow, which will affect the overall system efficiency,” says Xie, who is also helping the Massachusetts Department of Transportation develop a plan for connected vehicle technologies through research work with the UMass Transportation Center.

Xie’s research on intelligent transportation systems includes looking at how self-driving cars navigate four-way intersections. While some envision cars using vehicle-to-vehicle communication to weave around one another in intersections without stopping, Xie doesn’t think that will work. “First of all, it’s scary. You have to consider the passengers in the car,” he says. “And the other problem is, how do pedestrians and cyclists cross the street?”

Instead, Xie says a more sensible solution is vehicle-to-infrastructure communication, where traffic signals will know when cars are approaching from a distance and can adjust accordingly. “Even though your vehicle is still on the Lowell Connector, those traffic signals at the end of the off-ramp can be prepared,” he says. “Or, those controllers can guide your vehicle to a different route altogether.”

Enabling cars to communicate with traffic lights and roads will require some considerable infrastructure work, however. UMass Lowell is collaborating with five other universities from across New England to create the Transportation Infrastructure Durability Center. Backed by $2.5 million in funding from the U.S. DOT, one of the center’s goals is to develop new technology platforms for automated and connected vehicles that are durable enough to withstand harsh New England winters.

“We expect to develop new sensing and monitoring technologies and new construction materials to achieve a sustainable and durable transportation infrastructure in the U.S.,” says Tzu-Yang Yu, an associate professor of civil and environmental engineering who is heading up UML’s role in the project.

Start Your EngineS

Cuneyt Oge’s wife, plastics engineering alumna Margo Oge ’72, ’75, took her plug-in hybrid Chevy Volt to the car wash one day near their home in McLean, Va. As the former director of the Office of Transportation and Air Quality at the U.S. Environmental Protection Agency (EPA), Margo Oge helped craft the Obama administration’s landmark 2012 deal with auto-makers to double the fuel efficiency of cars and trucks sold in the U.S. by 2025 while cutting greenhouse gas emissions in half. At the car wash, she noticed a man waiting to pick up his freshly waxed Tesla. Oge was curious why he chose to drive an electric vehicle.

“Are you an environmentalist?” she asked.

“Absolutely not,” the man replied. “My previous car was a Ferrari, but this is more powerful.”

Oge, who wrote the book “Driving the Future: Combating Climate Change With Cleaner, Smarter Cars” after retiring from the EPA in 2012, says the car wash anecdote gives her reason to believe the future of driving is electric.

 “In the last seven years, the cost of the battery, the most expensive part of an electric vehicle, has come down by 70 percent. And all the car companies are investing in electric powertrains,” says Oge.

But if cars of the future drive autonomously with complex sensors and computer systems, then mechanical engineering Asst. Prof. Hunter Mack wonders if purely electric vehicles will be up to the task. He sees hybrid vehicles (with biofuel-pow-ered internal combustion engines that help carry the load) as the answer.

“Being able to run the lidar that tells you where all the different hazards are, or being able to run the communication systems, those require a lot of energy,” says Mack, whose research expertise is on fuels and combustion. “Our reliance on autonomous vehicles is going to extend the need for internal combustion vehicles in the long run, which is some-thing interesting that I didn’t see coming 10 years ago.”

Mack believes autonomous vehicles will have a “massive impact” on emissions and fuel efficiency, as they will eliminate many wasteful human driving habits (such as accelerating toward a stoplight before stopping). Engine design will also become more practical, eliminating cars that can accelerate to 120 mph in 10 seconds. “There’s no car that actually needs to do that,” he says. With the auto industry at the crossroads of autonomy and electrification, Mack says it’s an opportune moment for engineering innovation.

“For years, fuels and engines developed in their silos. But now, with all these external factors changing, we’re recognizing that we can optimize them in tandem,” he says. “Where we end up in five to 10 years is going to be very interesting.”

LICENSE AND REGISTRATION, PLEASE

The popularity of ride-sharing apps, combined with mobile technology that makes it easier to connect virtually, has led to an interesting trend: Fewer young people are getting driver’s licenses. According to a University of Michigan study, a record-low 71.5 percent of high school seniors had a license in 2015.

If you can order a self-driving car on demand to get you to the work or to the store, why would you want to buy a car? After all, it’s estimated that cars are parked and unused 95 percent of the time. Doesn’t it make more sense to share a car than to commit to car loans and insurance payments? 

“I get asked all the time if autonomous vehicles will signal the end of car ownership,” Evans says. “But car ownership is already a thing of the past in many countries. It’s just not replaced by autonomous vehicles, but [instead] by public transportation.”

Still, there were more than 17 million vehicles sold in the U.S. last year. Owning a car, truck or SUV remains a big part of our identity. Talay notes that Americans change their cars about every four years. “But if autonomous vehicles become the norm,” he says, “we will need far fewer cars on the road.”

Which is why automakers like GM, Ford and Toyota are getting in on the ride-sharing action. “They need to get into the ride-hailing business, which will be a lot more profitable,” says Talay, who also predicts that tech companies like Google and Apple will join forces with traditional automakers rather than try to build their own vehicles from the ground up.

Talay points out that 150 years ago, when steam-engine cars first rumbled on the scene, few thought they could replace horse-drawn carriages. Now you only see carriages at royal weddings and in places like New York’s Central Park. A century from now, will human-driven cars be relegated to a similar novelty existence?

"I believe they will coexist,” Talay says. “Some people will still drive cars, especially in rural areas and harsh climates. And some people will keep buying traditional cars anyway, just like some people still ride horses.”

Two UML alumni with a vested interest in the future of cars are John ’90 and Karen ’90 Manelas, who met at the university while earning their degrees in electrical engineering. In 2004, they both quit their corporate jobs to start their own car repair and service shop, Auto Care Plus. The business has grown to six locations across New Hampshire and Maine, and this fall, they’re opening a seventh in Derry, N.H., that will cater primarily to hybrid and electric vehicles.

“They’re building out the charging infrastructure pretty heavily in Europe, and it’s coming this way, too. I definitely see a big shift happening in the next 15 years,” says John Manelas, a self-described “motorhead” who used to build his own muscle cars while growing up in Lowell. He says one of his biggest challenges is finding mechanics who are up to speed with the computer components in new vehicles: “The technology seems to change every minute. It takes a certain mindset to work on cars today.”

Self-driving cars will only make things more complex with their wireless communications, cameras and lidar systems, creating entirely new avenues for services and products that few people have imagined.

"There are companies working on air fresheners for autonomous vehicles to try to keep people from getting carsick,” Manelas says. “That’s another challenge: People on their phone or laptop getting nauseous while their car drives them.”