Yale robot guru says no revolution likely

According to mechanical engineering professor John Morrell, the biggest challenges facing scientists have to do with the integration of multiple advanced technologies.
According to mechanical engineering professor John Morrell, the biggest challenges facing scientists have to do with the integration of multiple advanced technologies. Photo by Yale SEAS.

John Morrell, assistant professor of mechanical engineering and materials science, former director of systems engineering at Segway, and current director of the Center for Engineering Innovation and Design at Yale, studies and creates machines that connect hardware and software with human action. The News spoke with him about his research and current projects.

Q What is the current project that you are working on that you are most excited about?

A The current project — there’s a couple, because I tend to multitask — I’m interested in what we can do with computation to help people be more effective. That’s loosely called the field of human-machine interaction, but I tend to think about not so much human-computer interaction, which is interacting directly with the computer, but more how we interact with machines that may have computers in them, like cars, or transportation is where its most obvious right now, but robotic instruments, robotic surgery. So the project we’re working on right now … I have a student Jean Zheng who’s working on how we perceive touch in a peripheral way. We know that peripheral vision is a phenomenon where we can have our eyes focused on something and still perceive things in the periphery. The same is true of our sense of touch, but it’s not clear how it works, and so this notion of peripheral sensing is something we’re trying to explore and it’s pretty new. We don’t really know what we’re going to find from it. I find it really interesting because it’s a mix of both the design of the device that may be stimulating your sense of touch, physiology — human perception, and cognitive science, where we study what we pay attention to. I love that it’s cognitive science, physiology, and machine design all in one. That’s probably the one I’m most excited about.

Q What was the reasoning behind your tele-operated door-opening machine?

A In the world of robotics, if you look at where robot systems have been very useful in the past decade, it’s places like the Middle East, where you’ve got roadside bombs or dangerous environments … where you’d like to go in and get a look around, as well as humanitarian situations where you’ve got massive destruction, whether earthquakes

hurricanes or whatever. There are times when you would like to get a look around without having to put a person at risk. The robots that currently exist are track vehicles but they can’t even get through a door in most cases, or getting through is an extremely complex operation. So I was trying to figure out if there was a way to make a very simple device with fewer computers, fewer sensors, that could get through a door. Right now, if you’re a bad guy, a countermeasure against one of these robots is to just close the door. And the robots either have to blow the door up … and if you’ve got civilians behind it that’s no good. I was basically looking for not a disposable solution, but an expendable solution. Right now a lot of the times robots are so expensive that people don’t want to put them at risk. The other project that’s going on in the lab I’m pretty excited about is related to this one, which is trying to improve stair-climbing capability of balancing robots.

Q It sounds like your philosophy on robotics is humanizing technology.

A I believe that the big challenges we face are integration challenges — we have a lot of very advanced science. We have a lot of science discovery that we haven’t been able to realize productively in our human existence … and when we try to integrate lots of technologies together, we often increase the complexities of our lives rather than simplifying it. I think that’s a normal evolutionary process, but I like to be sensitive to the needs of human beings when we think about the technologies we put in our life. It’s part of my Yale education.

Q Do you have any technologies in mind that we haven’t realized?

A The stated promise of technology is still pretty far ahead of where we are, whether it’s in robotics or medicine — but less in medicine. Expectations have been far in front of what we can do for the past 20 years or so in the robotics field.

Q Can you predict a date for when robots take over the world?

A Well, I don’t think they will. I think robots are pretty impressive, but I think a lot of the people working on robots forget how impressive human beings are, looking at what human beings are capable of. I think automated systems are coming — obviously, they’re here. If you look at how much effort goes into checking in at the airport, they have a lot fewer people at the counters. You walk in, you swipe your credit card, it starts a conversation with a bunch of computers all over the place, it figures out who you are, probably checks in with Homeland Security, does a few other things, so they need fewer people there. That kind of movement of information is fairly easy. But, when you talk about manipulation, there still isn’t a robot that can pick fruit off a plant. There still isn’t a robot that can put a spoon to someone’s mouth and feed an elderly person. Manipulation is a really hard problem and we’re a long way from that. So computer devices can exert a lot of control and a computer network that’s run amok could take down the power grid and cause all kinds of problems, but we can still unplug them. They’re so far from being able to do basic manipulation tasks, which is why it’s an important research area. They’ll only take over the world if we abdicate all responsibility and let them. That’s probably the bigger risk: that we’re getting too passive. I believe in the value of using our bodies, so when people talk about using computers to save work over and over, I’m sort of thinking, do we really want to obsolete our bodies? Because there’s a lot of cells dedicated — our brain, some fraction of the cells in my body — are dedicated to actually being able to do something physical and to spend time trying to obsolete those cells seems like a mistake.

Q That was a really thoughtful answer.

A Well, I started looking at robotics sometime in college and then in grad school and I was really excited about it, but I also like using my body. I’m an athlete and I like making things and I really think it’s a mistake to obsolete our bodies. We’ll get to the right answer eventually, but we may have swung too far toward labor-saving at this point.

Comments