UMass Lowell student and faculty with NASA robot Valkyrie at the NERVE Center
Since Valkyrie arrived at UMass Lowell, the NERVE team has been painstakingly fine-tuning the algorithms that control her balance and motor functions, slowly teaching her to walk.

09/30/2016
By Ed Brennen

Maybe you’re an accountant, or a chemical engineer. Perhaps you take care of the elderly, or teach high school Spanish or manage an IT team. Whatever the field, imagine that you’re applying for a new position. Naturally, you speculate about your competition: Will they have more experience, better references, a lower price tag?

It’s less likely, presumably, that you wonder if the other job candidate’s brain is controlled by a computer.

But maybe you should.

Two-thirds of Americans polled by the Pew Research Center in 2015 said they believe robots and computers will do “much” of the work currently done by humans within 50 years. In many industries, it’s already happening. For decades, robots have been fixtures on factory assembly lines. Many of the tasks done by travel agents, bank tellers, cashiers, toll-takers and librarians—to name just a handful of fields—have been taken over in recent years by the cost-effective convenience of self-scanners, ATMs and the internet.

“The NERVE Center is the most incredible robot test-bed I’ve ever seen.” -Kimberly Hambuchen, NASA’s deputy manager for human robotics systems
Surgical robots are regularly used in operations on hard tissues (like bone), and a recent study at Children’s National Medical Center in Washington, D.C., revealed that a supervised autonomous robot could also successfully perform more challenging soft tissue surgery. In Japan, robot hotel workers are checking in guests and delivering room service; in China, robots are cooking and serving meals in restaurants; and on many Royal Caribbean cruises, they’ll mix you up a fresh margarita. Meanwhile, companies like Tesla, Google, Uber and Ford are developing “fully autonomous” cars that they say will be available to consumers by 2021.

And if UMass Lowell researchers have their way, robots will allow us to one day live on Mars.

Despite all this, most people like to believe their profession will always be immune to automation; 80 percent of the same respondents in the Pew study believe their own jobs will still exist in their current forms in 50 years.

After all, even Mars-bound robots like Valkyrie, NASA’s $2 million humanoid robot that graces the cover of this magazine and currently resides at the university’s New England Robotics Validation and Experimentation (NERVE) Center, could never possess the emotional intelligence, the intangible creativity or the all-around nimbleness that makes each of us so special—and indispensable—at work. Right?

Get to Know Valkyrie, Our Resident Humanoid Robot

Video by Alfonso Velasquez
The state-of-the-art NASA robot prototype Valkyrie arrives at UMass Lowell! She will stay at the New England Robotics Validation and Experimentation Center for the next two years.

TIME TO LEARN SOME NEW SKILLS

The good news is you probably don’t need to dust off your résumé just yet. But experts across the university agree that rapid, oncoming advances in artificial intelligence and machine learning will continue to revolutionize the role of robots in the workplace, changing life as we know it.

“It’s not that we won’t have jobs; we’ll just see a shift,” says computer science Prof. Holly Yanco, director of the NERVE Center and founder of the UMass Lowell Robotics Lab. “Look what’s happened with technology over the years. We don’t make buggy whips anymore, right? If robots are now doing jobs that people used to do, like picking and placing goods in an e-commerce situation, well, now we need people to build the robots and program the robots and take care of the robots.”

Computer science Assoc. Prof. Ben Liu sees three driving forces behind the robot revolution. The first is advances in machine-learning algorithms, the “brains” of a robot that are designed to adapt through an endless process of trial and error, teaching them how to navigate a world beyond traditional computer programming. The second is the explosion of big data from across the globe, and the third is exponential improvements in computation power to crunch all this data.

“Ten years ago, even with the best algorithm, we still were not able to run computers fast enough to learn from the data. Now, it’s hard to see where it’s going to hit the limit,” says Liu, who has little doubt that the resulting rise of robots in the workplace will have a major impact on society. “It will replace more and more cognitive, white-collar work, that’s for sure. The challenge for us will be to adapt and learn new skills.”

With the confluence of those factors, there’s never been a better time to work in robotics. And, with apologies to Silicon Valley, there’s no better place to do it than right here in Massachusetts, where more than 150 robotics companies currently employ around 3,200 of the area’s best and brightest computer science and engineering minds. Conveniently enough for UMass Lowell, the three biggest robotics companies (in terms of number of employees) are all located within a 15-minute drive from campus: Brooks Automation in Chelmsford, Amazon Robotics in North Reading and iRobot in Bedford.

UMass Lowell Prof. Holly Yanco stands with robot, Valkyrie
Holly Yanco, computer science Prof., director of the NERVE Center and founder of the UMass Lowell Robotics Lab stands with Valkyrie.

“We’re developing a real robot ecosystem in the area,” says Yanco, co-chair of the Mass Technology Leadership Council Robotics Cluster, which was formed in 2005 to spur industry growth through knowledge-sharing and relationship-building among the region’s leading robotic companies, academic research labs and policy-makers. Rather than operate in silos in the race to build a better robot, the Robotics Cluster encourages companies to take advantage of the area’s wealth of world-class universities and R&D resources.

Front and center is the NERVE Center, which is one of only three robot test facilities in the country with courses designed in conjunction with the National Institute of Standards and Technology. Inside the 10,000-square-foot facility located just two miles from the campus center, developers can take their robots for a spin on a variety of terrains and inclines, run them through obstacles and water, and test their manipulation capabilities, receiving the all-important third-party validation needed before taking them to market. The collaborative efforts are paying off. Led by the likes of iRobot’s Roomba and Amazon Robotics’ warehouse fulfillment robots, sales of Massachusetts-made robots are approaching $2 billion annually—and growing. The global robotics industry, already a $71 billion market in 2015, is expected to nearly double to $135 billion by 2019, according to tech research firm IDC.

Boosting UMass Lowell’s stock in the robotics industry is Yanco, who is widely recognized as a leader in the field. “Holly is very plugged in,” says Vice Chancellor for Research and Innovation Julie Chen, “and I’ve learned in talking with her that robotics is a very close-knit community. Everybody is separated by two degrees.” Kimberly Hambuchen, NASA’s deputy manager for human robotics systems, is excited that Yanco’s team has been entrusted with Valkyrie.

“The NERVE Center is the most incredible robot test-bed area I’ve ever seen, and Holly is one of the leading experts in the field of human-robot interfaces, so I’m really excited to see what sort of work Holly and her team can bring to the table,” says Hambuchen, who visited campus in September to deliver a guest lecture called Exploring Space with Robots: Do We Need People? “We’re expecting to get a lot of good things out of them, things that won’t be replicated with the other university robots.”

BIG BOT ON CAMPUS

Most robots on the market today are single-purpose, such as the industrial robot arms that are programmed to perform repetitive tasks on assembly lines. Service robots, like Roomba vacuums and bomb-disposal units used by law enforcement, are another type growing in popularity, designed to do jobs deemed too dull or dangerous for humans. Then there are the more complex collaborative robots, or “co-bots,” which are designed to work autonomously alongside humans in a variety of settings, from offices to hospitals to stores to farms.

Or, in the case of Valkyrie, on Mars.

Developed by NASA’s Robonaut program, Valkyrie has become a campus celebrity since arriving at the NERVE Center in April, drawing the attention of all of the major Boston news outlets, as well as nationally from the Associated Press, the History Channel and National Geographic. One look at the 6-foot-2, 300-pound R5 (as Valkyrie is technically known) and it’s easy to understand the fascination. With a gloss-white frame highlighted by gold trim, a sleek black visor and helmet-mounted lights and sensors, Valkyrie looks like a life-size “Star Wars” action figure. Of course, NASA has much bigger plans for her. The space agency hopes she’s the prototype for a next-generation version that will travel to Mars and pave the way for human exploration—all within the next 25 years.

But before Valkyrie version 2.0 can set foot on the Red Planet, Valkyrie version 1.0 must learn how to walk in Lowell.

Robots Around the World

“It’s amazing to have Valkyrie here,” says Yanco, who is leading a team of more than a dozen student researchers from UMass Lowell, in collaboration with Northeastern University, over the next two years to teach Valkyrie how to walk, manipulate objects and understand the world around her. “She falls sometimes, but we’re learning to program her to walk better and balance better.”

Valkyrie is actually a quadruplet—her three sisters reside at MIT, the University of Edinburgh in Scotland and NASA’s Johnson Space Center in Houston. Chen says that having one of four Valkyries in the world here on campus is a coup for the university. “Everyone can appreciate going to Mars,” she says. “It’s helpful in getting people to appreciate and understand science and technology.”

Since Valkyrie arrived on campus, the NERVE team has been painstakingly fine-tuning the algorithms that control her balance and motor functions, slowly teaching her to walk.

“It definitely makes me appreciate the human machine,” says Jordan Allspaw, an undergraduate computer science major who spent two weeks in January at the Johnson Space Center learning how to train and maintain an R5. 

Computer science graduate student Carlos Ibarra Lopez, who in 2013 was part of the UMass Lowell “Rover Hawk” team that won NASA’s Robo-Ops competition in Houston, says that working on a groundbreaking project like Valkyrie is as thrilling as it is challenging.

“The idea that anything we’re working on would end up on Mars is pretty amazing,” says Ibarra Lopez, who also trained in Houston in January. “But since it hasn’t been done before, you can’t really just Google something to find how to solve problems. We have to contact NASA.”

While her NU counterparts, professors Taskin Padir and Robert Platt, are primarily focused on Valkyrie’s balance, movement and ability to grasp objects, Yanco is busy analyzing how the robot will interact with humans. Valkyrie can’t talk (yet), and although her black facemask may serve her well in a poker game, it doesn’t really let her convey emotions. For now, she relies on a glowing NASA emblem on her chest plate to communicate some basic concepts and interact with those around her.

“There’s a very broad range of human-robot interaction, which I find very interesting to think about,” Yanco says. “For less-capable robots, the interaction could be using a joystick to control a bomb disposal robot. If you’re talking Valkyrie on Mars, the interaction is a little different. It might have to send messages back to NASA saying it’s having trouble.

One of my grad students is running a survey on the internet to see what icons people could look at and say, ‘I understand it means the robot needs help.’”

And then there’s the question of how Valkyrie, or any autonomous robot, can best communicate with people around them who aren’t familiar with them. Yanco sees the impending proliferation of self-driving cars as the most obvious example of this problem.

“When we cross the street we make eye contact with the driver of a car. What do you do with an autonomous car? There’s no one to make eye contact with. How do you know it’s safe and they recognize that you’re there?” she says.

“As more and more autonomous systems are deployed, we need to understand what they’re doing. Are they moving? Are they not moving? Can we get near them? Should we not get near them?”

In addition to studying this human-robot interaction and validating the autonomous skills that Valkyrie would need for deep space missions, Yanco and her team are helping NASA develop test courses for the 2017 Space Robotics Challenge, in which teams from around the country will control a virtual Valkyrie through a series of tasks.

“There aren’t many other places in the world where I could be working with a robot like this,” Lopez says.


Socio-Economic Implications

Earlier this year, researchers and academics, including Sociology Department Chair and Assoc. Prof. Mignon Duffy, gathered at Mount Holyoke College for a conference on “The Future of Jobs: The Dual Challenges of Globalization and Robotization.” Duffy’s research focuses on care work such as health care, child care and education, one of the fastest growing sectors of the labor market.

“It’s also one of the most resistant to being fully automated, given the need for human responsiveness and relationships,” says Duffy, who is also associate director for the Center for Women and Work.

“The conference was specifically focused on work and workers, but I do think that is one of the major things we need to think about with the rise of artificial intelligence: How does this impact jobs as we know them? It has the potential to have a huge impact.”

While Duffy believes care jobs require “invisible pieces” such as human relationships and emotional responsiveness that even the most advanced robots just can’t imitate, she is concerned that the rise of artificial intelligence could change the way society thinks about those whose job it is to care for others.

“Since we often do not recognize and value the important emotional and relational work that people like teachers and nurses do, could we then try to reduce their jobs to a series of tasks that could be automated?” she says. “In some ways automation is the next stage of standardization; if we think something can be standardized, broken down into a series of tasks, then it can be automated.”

Looking 50 years down the road, Scott Latham, associate professor of management in the Manning School of Business, sees workplace automation creating an even wider socio-economic divide than exists today.

“There will be no more minimum-wage jobs. Robots will take over a lot of those jobs—everything from landscaping to working in restaurants to home care,” says Latham, who researches business strategy. “The jobs will be in industries like robotics and biotechnology, where higher knowledge workers are required. Unfortunately, that means there will be an even greater disparity in wealth.”

To counteract this disparity, Liu says that the workforce will simply need to adapt and keep learning new skills, just as it has always done when technology evolves.

“I think we’re in a unique position at UMass Lowell,” Liu says, “because of our technology advances and lifelong learning and job training, to provide these opportunities.”