RoboHub speaks with Brian Bourke, 570 News

Wednesday, October 31, 2018

To celebrate the opening of the Engineering 7 building our manager, Brandon DeHart, and Professor Sanjeev Bedi went on Kitchener Today with Brian Bourke to discuss the RoboHub and the new Ideas Clinic facility. The full show can be listened to online beginning at timecode 44:14 or the transcript of the interview is featured below.

Brian Bourke: There are, as we’ve mentioned before a lot of people, a whole lot smarter than you and me and a lot of other people in this town. Its why there are so many interesting things going on. It’s why people have made glasses where the internet comes up a couple of feet from your face and all those kind of odd things that you discover about this city. It was interesting earlier this week when the very sexily named engineering building 7 opened up at the University of Waterloo and along with that RoboHub! I am joined by – and this might take awhile – Sanjeev Bedi, Founder and Director of the Engineering Ideas Clinic and NSERC Chair in Immersive Design Engineering Activities (IDEAs) and a Professor in the Department of Mechanical and Mechatronics Engineering at the University of Waterloo; and Brandon DeHart, the Manager of RoboHub. Those are two very distinctly different titles. Thank you, gentlemen, for joining me this afternoon.

Sanjeev Bedi: Thank you, Brian, for having us.

Brandon DeHart: Thanks for having us.

Brian: So the opening of Engineering building 7, that’s not the sexiest name in the world guys.

All: [Laugh]

Sanjeev: You think of sexy right, you can’t depend on engineers to come up with sexy.

Brian: My brother’s an engineer, so yeah, I fully understand what you’re saying.

Sanjeev: Faithful is what we aim for.

Brian: What does this building mean to the University of Waterloo, right now?

Sanjeev: Well, the key thing is that there are many different buildings that pop up in different schools, but the purpose of this building is what distinguishes it from every other building. So the two floors - the first two floors - are basically dedicated to student activities, that’s a departure from everything that you can think of. See at Waterloo we are very, very fortunate to have some of the best students in the country coming here; when they come over here they have a certain purpose in mind, they want to become engineers! And when you talk to experts, experts who say “hey we have to build a foundation for these students before they can do any engineering”, so for the first two years or three years we basically give them foundational courses in mathematics, physics, chemistry, and by the time the students reach third year, they forget what they were here for!

Brian: [Laughs]

Sanjeev: So then their sole aim becomes to graduate from here and sort of go hide behind a computer somewhere; and they do well, they are very brilliant kids, so that’s not a problem. But what we have decided is that we want to make professionals that will be productive as soon as they step out of the university. What we have done is, we are basically making sure that the students indulge in some engineering activity as they go through the foundational knowledge formation. So while the students are studying we are also doing some engineering based activity. [For example] On Monday when the building was inaugurated we had civil engineering students, they were given a two day period; they basically designed a road, they found that the road has a bridge on it, they actually designed the bridge on paper first, then they built the bridge, and finally the whole activity culminated with them testing the bridge. Now, this is not a bridge that you would put over the Grand River, or anything like that, but the rush, the concepts, the thinking, the thought process, that is very similar to what a real engineer would do faced with building a bridge over the Grand River. This what it does is, it ignites the reason why they are here and gives them context to the learning that they do in our classrooms. So, it was really well received, the students although they had an exam on Tuesday, they stayed late and continued doing the bridge, so I actually had to ask them to leave just because they wouldn’t have left otherwise! [Laughs]

Brian: So these students who come in, do they come in with an idea of what they want to do, what they want to build, what they want to create? Or is that part of the discovery as they are at UW (University Of Waterloo)?

Sanjeev: Very good question Brian! See in grade 12, the students have really not seen too much of the world; they basically rely on trusted people to give them advice, typically parents, teachers in high school, uncles, aunts and their friends, right? They have a notion of what engineering is all about, they’ve actually not done much of engineering, so what we have to do is, we have to give them a taste of different flavours of engineering within a particular program, to ensure that they select what they really want to do for the rest of their life.

Brian: I would imagine that there are some who come in there who completely turn direction from what they maybe thought they wanted to do and what they end up doing two or three years later, right?

Sanjeev: Absolutely, I’ve tracked a couple of students they came here they say “my passion is mechanical engineering!, a little way later they turn around in second year or so,  and they say “ no, no, no… I want to do a little bit of mechatronics” and finally settling in for a mix between mechanical engineering, electronics and software. So the bottom line is, at Waterloo because of the co-op program, we give the student an opportunity to work in six different environments, in six different companies, for six different bosses. This immersion in six different environments is what builds the foundation for innovation in the future, it is what builds their design vocabulary, [and] it is what builds their repertoire of things to draw from as new challenges are put in front of them during their career. So yes, they do change direction, but I think they build on the aspirations that they came to realize in our departments in our programs.

Brian: So it was fascinating earlier this week to see the work of the robot TALOS, I believe his name is, which sounds like a Marvel Comics villain by the way [laughs]. So it was fascinating to see that, what is the point of RoboHub?

Brandon: Just to catch the sideline there, TALOS was actually a mythological figure from Greek mythology, it’s been used in a number of different places there’s companies, there’s organisations, there’s models, etc. This particular model, much like you would say, you know Honda Civic, this is a PAL Robotics TALOS. So we actually have a naming competition running right now, to figure out what we want to name our TALOS. The RoboHub itself is all about kind of mixing collaboration and robotics at every level. So it’s bringing robots together, different robots, so whether it’s a drone and a wheeled platform and something with legs and having them work together, robot-robot teams, robots with humans, working with humans. And then also getting to the point of academics collaborating, academics working with industry at every level of robotics. So that we can build this core hub, RoboHub, of robotics and collaboration at every level.

Brian: So Brandon DeHart, where are we in the world of robots right now? Because I think most people are familiar with them, we can see them on an assembly line, and that’s kind of as far as we’ve gone. And then we look at the science fiction level of robots. Where are we? What’s the future looking like for the next couple of years for robotics?

Brandon: I would say over the next few years there’s going to be a lot more on the collaborative side. So we’ve seen, as you’ve said, the robot arms on the assembly line and they’re in a cage they’re away from people. Most of them can’t even sense if or when people are nearby, so they have an additional layer of security that shuts the robot down if anybody gets close; things like that. And a lot of the developments we are seeing now, related to cyber-physical systems and industry 4.0, brings in this collaborative element so that the robots are actually working side by side with the humans; they’re doing the pieces of the human's task that are dull or dirty or dangerous, whatever it might be. And kind of augmenting their abilities, whether that physical, or parts sorting or image recognition whatever it might be that the humans aren’t as good at and the robots are better at so together we can actually achieve more productivity, more effectiveness etc.

Brian: So is it legitimate to be concerned, that at some levels, eventually, the robots will take over?

Brandon: [Laughs] I don’t think we need to worry about that. You know we're hearing a lot in the news about how artificial intelligence is going to allow things to take over and get away from us and all these things, but at the moment, and as has been the case for decades, artificial intelligence for the vast majority of purposes, is really just a pattern recognition machine. So if you give it hundreds, thousands, millions of examples of something than it can learn whatever that might be. But there is no, basically no progress being made right now, or only very little, at kind of the fundamental exploratory level into generalized intelligence, creativity, curiosity, inventing new things or responding to a situation that has never happened before, which you and me do every single day.

Brian: Is there a point, maybe a few years down the road, where we begin to, as a general public, interact with the robot world to some degree?

Brandon: I would say we already are, and it really depends on what your definition of robot is. Personally, my definition [of a robot] has always included embodiment, so there has to be a physical aspect to it, where it can sense its environment and then change its environment, that’s kind of my definition of a robot. Many people have much broader scale things, where an algorithm running on a server somewhere could be defined as a robot, depending on what its inputs and outputs are. But even something like your car, if it has lane assist or intelligent cruise control or anything like that, is effectively sensing its environment and then changing its environment, by moving at a different speed, different rate, etc. The one thing I like to pull, that I learned here at Waterloo as I was going through undergrad, it’s only called a robot until we have a defined use for it, then we call it a car or a dishwasher, or whatever its actual job ends up being.

Brian: [Laughs] I hadn’t actually thought about that, but yeah that’s very true, unless your definition of a robot is like something that like you said, has an embodiment of somewhat human form and moves on its own, all those things are just responding right?

Brandon: Yes, they are all taking in input, they're all changing either internally or externally their environment. So really depending on your definition of robot, all of those things are.

Brian: Where does the work of a place like RoboHub really have the potential to make a difference for people?

Brandon: I think this is kind of a critical time for multi-agent systems, whether that’s robots working together or robots working with people. In places where, right now, the human might have to put on layers and layers of protective clothing and security and carry air in with them potentially, underwater, volcano, space, all these kind of things; where we can have a group of heterogeneous robots, different types of robots working together where there’s one on tracks that can maybe have a drone attached and it can use the drone to get a better look at something, the drone takes off, uses its camera, shares its map and then goes back and lands back on the roller. Or there’s a humanoid that’s going to go into an oil and gas plant or a nuclear disaster like Fukushima and can navigate the human-centric environments that we’ve been building all over the world, and beyond the world now. They’re really good for our form, hands, two legs, you know, a set of stairs with a handrail, all those things, and then the valves, all of the safety gear is built for humans. So if we have now decided or discovered that things are too dangerous for a human to do; potentially a single human, at an oil and gas plant, for example, could have a fleet of humanoids, like TALOS which has been built to, in the near future become an industrial humanoid, they could have a fleet of them and send them to various checkpoints, and then when they get to the “hard part”, the human can then take over and teleoperate and do that piece. So it’s effectively a way of removing the humans from danger, removing them from something like asbestos removal [which] potentially could be automated. You can have people remotely controlling robots that can handle all of the tools that we’ve built for ourselves to do these jobs.

Brian: Do you think there’s still a fair amount of resistance to the robotic world?

Brandon: I think most of the resistance comes from assumptions and a lot of it comes from seeing how many of the other industrial revolutions have taken place. One thing that I have believed, basically since I have started thinking about it, is the next industrial revolution that we're kind of on the cusp of right now with autonomous vehicles, and robotics and AI [artificial intelligence] kind of coming to their fruition now; is that it’s going to be a very different beast. It’s going to be a lot more about augmenting humanity, about a single independent person being able to do multiple jobs, potentially at the same time, things like that. Because we're seeing widespread unemployment and we're not able to fill jobs in manufacturing and assembly, in lots of industries that people are choosing not to do. And if we want those things to continue, because they are enabling the rest of society as it functions now, we need to have someone or something that is willing and able to do that. So if removing the person from the physical location helps that, then that’s a perfect way to go.

Brian: So let’s talk about TALOS again, you guys are looking to rename this robot. How do people go about doing that?

Brandon: So we're running a naming competition, it’ll run for the next few weeks. All of the rules are available on the Waterloo RoboHub website, which is uwaterloo.ca/robohub and you can tweet at us @uwrobohub or email robohub@uwaterloo.ca, all of those things. If you want to send in your name ideas and then near the end of the month for the last week of November, we'll have a list of our favourites of the list that have been sent in which you can then vote on and at the end of the month we'll have the winner, and the winner will be able to come in and actually meet TALOS in person.

Brian: Well I'm hoping that we'll be able to see some of those because I know that people have a very twisted sense of humour [laughs] and I know that engineers have a kind of twisted sense of humour, too, so I'm hoping we'll be able to see some of those as you get a little closer. This is really interesting stuff and its fascinating to see where we are going and you guys bring a completely different perspective, so congratulations on the new building and congratulations on RoboHub and thanks for joining us!

Sanjeev and Brandon: Thank you, Brian!

Brian: Sanjeev Bedi and Brandon DeHart, both from the University of Waterloo, Brandon is the Manager of RoboHub while Sanjeev has a really long title, but let's just know that he is a Professor in the Department of Mechanical and Mechatronics Engineering and the Director of the Engineering Ideas Clinic, so he too working with a lot of people smarter than us. It's kind of cool. I can think of some pretty interesting ideas for robots. This is Kitchener Today on 570 News.