Grant Imahara Generation Robot

Grant Imahara on the 'Wild West' future of robotics

Contributed by
May 29, 2018

Former MythBusters and Battlebots wizard Grant Imahara certainly knows his way around a robot. That's why Mouser Electronics has the engineer doing a new video series specifically focused on the people who will grow up knowing robots their entire lives and the kind of cutting-edge tech they'll be using.

That's the subject of Generation Robot, Imahara's five-part series in which industry experts explain the way we use robots now and how that will change as we embrace new technology as a daily part of life. In the first video, the only one released as of this posting, the host travels to the Institute for Robotics and Intelligent Machines at Georgia Tech to meet with executive director Dr. Magnus Egerstedt. The second episode premieres on May 23.

Seeking our own dose of Imahara's insights and hoping to understand which pieces of robotic science fiction are becoming science fact, SYFY WIRE got to sit down with him to talk ethics, giant robot arms, and why robots won't be stealing astronauts' thunder any time soon.

Let's start with a hard one. What is your personal definition of a robot?

I tend to go with the more traditional one. A computer-controlled mechanism — not specifically a being, though in many cases it is — that responds to stimulus and makes its own choices as to the outcome. A robot should be able to make decisions.

You say in the first Generation Robot video that a robot playing the marimba is one of the more impressive that you've seen — is that a good example?

Exactly. It's capable of improvising on its own, which is incredibly cool. Jazz is complicated. A lot of it is unstructured. During the process of this robot's improvisation, you can see that it has to communicate with the other musicians.

In order to respond to the cues, they took a dexterous robot arm and put a little head on it that bobs up and down along with the beat. Much like the visual cues the human musicians use with each other, this will look over at the other musicians and take their cues. Like human musicians, it's looking and asking "is there something I can do that will fit into this?" The marimba-playing robot doesn't play in a vacuum. It's not playing a pre-programmed sequence with a few deviations. It has to look at, give cues to, and take cues from the humans in the group.

In terms of them needing to understand people, is finding these channels of communication the next step of innovation for robots?

Certainly one of them. And it'll be key to integrating robots into our daily lives. Human communication above and beyond the words that we say is so nuanced. It makes it difficult to not only analyze the vocabulary you use but the intention behind it. That's something even humans have difficulty doing, let alone a robot.

So do you think other disciplines will feed into engineering to solve these problems?

Absolutely.

A lot of the emphasis is on how to get robots to recognize human emotion. Engineers are great at many things. They're not necessarily great at recognizing basic human psychology. We as engineers can't do everything. We can't exist in a vacuum either. We're going to need help from unexpected places.

When you're trying to implement robots into our daily lives, into the workplace, that requires the ability to deal with situations that may come up that you didn't expect. That's the nature of dealing with humans.

Do we have to design robots that people want to interact with? The big arm that feeds you yogurt in the first video would send a kid screaming if you sat one down in front of it.

One of the biggest problems facing personal robots, robots in our homes, is finding a design that's pleasing to everyone. Pleasing to kids, pleasing to adults, pleasing to elderly people. One of the areas of emphasis is the "realistic human" and that one is very difficult because of the Uncanny Valley. No one has done it yet. The other side is take something, like a face, and make it as non-human as possible. Look at [Big Hero 6's] Baymax, for example. He's an animated character, but as an example of product design, he's got big eyes and he's non-threatening.

The other big factor, when you consider something like that big robot arm, is you have to consider what's available. That robot arm is connected to a robot called PR2 and the idea behind having a giant — which you don't need to feed you yogurt — is that you want the flexibility to be able to, say, help someone out of bed or get dressed. When you're designing a robot helper, you want them to do all kinds of things. Not "this is just a robot that feeds me yogurt? It doesn't do anything else?" You want that robot to do as many jobs as possible. So when you combine that need with what motors are available, you get bulky-looking things.

Isaac Asimov listed three laws of robotics in I, Robot. Is there a law agreed upon in the robotics community for designing robots for the home?

Asimov's laws are an ideal standard. Currently, it feels like the Wild West, where everyone's racing towards creating something for the home and I don't think there's any regulating body for how they perform.

If we can get the spirit of Asimov's laws, I think that's the best that we can hope for.

[Generation Robot's second episode will focus on ethics]

There's a growing push to go back to the Moon and use that as a jumping-off point for Mars — that means very few spacewalks and doing lots of it through robotics. A robotic geologist was just launched to Mars a few days ago. Does this shift the relationship between the human trailblazer and a robotic one?

When you think about exploring deep space, it's very dangerous for humans. To potentially sacrifice human life is a delicate proposition. On the flip side, if you want to send a robot to another planet for an extended period of time and at some point you know it'll cease to function, it's not considered a tragedy. The prospect of going to other planets becomes easier when you know you can send a robot to essentially do the dirty work. Even though we're sending the robots ahead, they're simply enabling us to move forward.

What is the main thing you want people to take away from the Generation Robot series?

I hope people can envision a future where we as humans have robots that are alongside us, assisting in our daily lives. Change is very difficult for some people. Being able to see what people are working on now, to enable the robots of the future to be our helpers and our companions will fill them more confidence and hope for the technologies to come.

This interview has been edited and condensed.