Latest / Automated with Brian Heater / How Robots Fail, Learn, and Improve: A Conversation with Holly Yanco

Transcript

How Robots Fail, Learn, and Improve: A Conversation with Holly Yanco · Automated with Brian Heater

0:00
SPEAKER_00: The use cases for telepresence robots where I think it could be really valuable is possibilities for allowing people with disabilities to travel to museums through telepresence robots.
0:09
SPEAKER_00: So if they couldn't get there, they could be there.
0:12
SPEAKER_00: I think there are good use cases.
0:13
SPEAKER_00: I know Vigo back in the day was working with letting students go to school using them.
0:18
SPEAKER_00: So if students were immunocompromised, they could use them.
0:20
SPEAKER_00: So I think there are cases for them, but I think we got so used to using Zoom and other video conferencing systems that, you know, do we need to have a presence somewhere else?
0:41
SPEAKER_01: Welcome to Automated.
0:43
SPEAKER_01: My name is Brian Heater.
0:44
SPEAKER_01: I am managing editor at A3.
0:47
SPEAKER_01: We have another robotics legend on the show this week.
0:51
SPEAKER_01: You'd think I would be sick of saying that by now.
0:54
SPEAKER_01: Uh, among a lot of other things, Holly Yanko is the Distinguished University Professor of Computer Science at the University of Massachusetts Lowell.
1:03
SPEAKER_01: Buckle up because you are going to learn a lot from this one.
1:08
SPEAKER_01: Thanks to Professor Yanko for chatting with us.
1:11
SPEAKER_01: Thanks to you for tuning in.
1:12
SPEAKER_01: Please like and subscribe and follow the automated newsletter over at automated.fm.
1:19
SPEAKER_01: Here's Professor Yanko.
1:21
SPEAKER_01: Professor Yanko, thank you for joining us.
1:24
SPEAKER_01: Uh, and congratulations on the upcoming move.
1:26
SPEAKER_00: Thank you.
1:27
SPEAKER_01: As we're recording this in September.
1:29
SPEAKER_01: One of the first comments I saw on your LinkedIn page is what's going to happen with all the Pez dispensers.
1:35
SPEAKER_01: And that must be a huge logistical part of moving from one city to another.
1:40
SPEAKER_00: Yeah, there's there's quite a few of them in my office.
1:42
SPEAKER_00: Uh oh, robots, of course, need to move, but uh yes, I have uh thousands of Pez dispensers in my office.
1:48
SPEAKER_01: Tell me a little bit about how you accrued those over the years.
1:51
SPEAKER_00: So this collection started, and here's some of them.
1:53
SPEAKER_00: Um back in 1993, uh, a group of us from grad school had gone away for spring break, and we went into a store and they had Tweety Bird Pez dispensers, which at the time, if you were watching Seinfeld, there was an episode of the Tweety Bird Pez dispenser.
2:08
SPEAKER_00: So somebody bought them for all of us and it ended up on my desk.
2:11
SPEAKER_00: And then people started bringing them to me.
2:14
SPEAKER_00: So I went from having one to ten to a hundred pretty quickly.
2:16
SPEAKER_00: And now nobody can bring them to me because they don't know what I have.
2:19
SPEAKER_00: So I have to buy all the new ones myself.
2:21
SPEAKER_00: But uh I've collected them over the last 30 years, and I think uh when students come to my office, they're perhaps a little overwhelmed.
2:28
SPEAKER_00: But but it's I think it's fun to have them, makes the office very colorful.
2:32
SPEAKER_01: It's nice to have a thing.
2:33
SPEAKER_01: I went to a school in at UC Santa Cruz and I visited a professor who has uh uh students from all over the uh travel all over the world and bring wing chimes back.
2:41
SPEAKER_01: So he's just got like a room full of wind chimes.
2:44
SPEAKER_01: And I was watching a presentation that you were giving on on Teleop, and you mentioned that the uh one of the robots had a Hawaiian shirt and we're debating whether or not that was a marked thing first or a robot thing first.
2:55
SPEAKER_01: It's always good in academia to have some sort of uh, I don't know, some kind of calling card.
3:00
SPEAKER_00: Yeah, and actually um when my PhD students graduate, so you may know that at Brown, when the graphics students graduate, they get a rubber chicken.
3:08
SPEAKER_00: Um, so I decided it'd be nice to have something to give my students.
3:11
SPEAKER_00: So I actually gave them a giant robot pez dispenser.
3:13
SPEAKER_00: So instead of dispensing individual PES, it dispense packs of PES, but it's a big robot pez dispenser.
3:19
SPEAKER_01: Yeah, and I promise we'll move on from the background after this, but I have to ask you how how you got your hands on a gold key pod.
3:25
SPEAKER_00: It was actually painted.
3:27
SPEAKER_00: Uh Wendy Ju had painted them for HRI 2012.
3:31
SPEAKER_00: So they were awards for the the video competition for HRI.
3:34
SPEAKER_00: And I was one of the general chairs that year.
3:36
SPEAKER_00: So she had made some extras um and she didn't want to take them home with her.
3:40
SPEAKER_00: So I ended up with it, it's just spray painted, but it looks really great.
3:42
SPEAKER_00: She did a good job with it.
3:44
SPEAKER_01: So tell me a little bit about this upcoming move.
3:46
SPEAKER_01: Um, the why and why now?
3:49
SPEAKER_00: I have been at the University of Massachusetts Lowell for the last 24 years, and it's been a fantastic place to work.
3:56
SPEAKER_00: And while I've been here 13 years ago, we started the New England Robotics Validation Experimentation Center.
4:02
SPEAKER_00: We call it NERV for short.
4:03
SPEAKER_00: It's a lot easier to say.
4:04
SPEAKER_00: I got the opportunity to move to UMass Amherst, which if you're outside the state, it's like, oh, she's just moving UMass systems.
4:11
SPEAKER_00: And I am, uh it's about an hour and a half away.
4:13
SPEAKER_00: But they asked if I would come out and help them grow their robotics program.
4:17
SPEAKER_00: And so that was a pretty exciting opportunity to do something new.
4:21
SPEAKER_00: And actually, what I think I like best about this move is that I'm still going to have the opportunity to to work with UMass Lowell.
4:28
SPEAKER_00: So the idea is that we will expand the NERV Center to be in Lowell and in Amherst.
4:33
SPEAKER_00: So here in Lowell, we've been testing exoskeletons, search and rescue robots, a lot of robot manipulation and a few other things.
4:41
SPEAKER_00: But what I'm interested in picking up when we go to Amherst being a land grant school is thinking about agricultural robots, particularly kind of small family farms.
4:49
SPEAKER_00: We don't in Massachusetts have kind of the large Midwest farms, so that kind of John Deere, but that there's a lot of work that the university does with the Extension School with family farms.
4:58
SPEAKER_00: And so can we think about what we could do with robot systems and how we could test them to help small farmers use robots?
5:05
SPEAKER_00: The other thing I'm thinking about expanding into is uh underwater robotics.
5:08
SPEAKER_00: We've been doing some underwater robotics here, but uh UMass Amherst, if you know the geography of Massachusetts at all, uh they have a site, uh, they have a marine station in Gloucester, Massachusetts, which is actually much closer to Lowell than it is to Amherst.
5:21
SPEAKER_00: But I want to take advantage of the fact that they have that marine station so that we can actually do some underwater work right off of the appearance.
5:28
SPEAKER_01: What's the through line through all of these modalities?
5:31
SPEAKER_01: Like what connects all of these different kinds of robots for you?
5:34
SPEAKER_00: So from my research perspective, I do human robot interaction.
5:38
SPEAKER_00: And I always say that for me, HRI is the same as robotics.
5:42
SPEAKER_00: And then because I have to give it to people, I'll say, well, it's robotics minus epsilon.
5:46
SPEAKER_00: Because you have to give people that little bit.
5:48
SPEAKER_00: Anything that we do to make a robot better, whether it's better path planning, um, whether it's better teleop, whether it's better it can be social.
5:56
SPEAKER_00: I don't do a lot of social robotics.
5:58
SPEAKER_00: I always joke that I'm an antisocial roboticist.
6:00
SPEAKER_00: But I think that all the things that we do to make robots better also improves how it interacts with people.
6:05
SPEAKER_00: People may not, it may not be the primary interaction that we're thinking, they'll work with it, but there'll be bystanders to it.
6:10
SPEAKER_00: So everything that we do to make a robot more understandable to work better is HRI.
6:14
SPEAKER_00: So that's I I think a lot about just how do we build great robots.
6:19
SPEAKER_00: But then the other part of me thinks a lot about how do I break robots?
6:23
SPEAKER_00: Um so this is my mug I'm using right now.
6:26
SPEAKER_00: Um this is our nerve center mug.
6:28
SPEAKER_00: And so edge cases are really interesting to me, right?
6:33
SPEAKER_00: So I think, you know, when Google first came out with their automated car, they would say, oh, we drove X hundred thousand miles, whether it was 200,000, then 500,000.
6:42
SPEAKER_00: And that's great.
6:43
SPEAKER_00: But if you think that they're doing most of their driving in kind of the Bay Area, that's very different driving than if you're in, you know, the mountains of Colorado or if you're in New England, where we don't bother to paint the roads most of the time.
6:56
SPEAKER_00: Uh, so so how do you how do you actually trust that what you're doing for that testing?
7:00
SPEAKER_00: Maybe it's really, really great in mountain view, but is it going to be great everywhere else?
7:05
SPEAKER_00: So, how do we think about testing robots?
7:09
SPEAKER_00: You want to test them quickly, but to hit as many edge cases as you can.
7:12
SPEAKER_00: So I kind of have this dual side of my world of thinking about how do we build better robots and then then how do I break them?
7:20
SPEAKER_01: It's a little tug-in-cheek, obviously, and you have you have a mug, but you do feel like to a certain extent it is your your role or your job to push robots to the point of failure.
7:30
SPEAKER_00: If we can't find where the failure points are in the lab, we're gonna find out where they are once we deploy them.
7:35
SPEAKER_00: So so we're better off to find them before we go for deployment.
7:38
SPEAKER_00: And and whether that's you know, exoskeletons or robot wheelchairs or underwater robots, I think that we we need to be testing them in a in a wide variety of scenarios, but we need to also do it in a finite amount of time because we want to also deploy our systems.
7:54
SPEAKER_01: Yeah, it's funny.
7:55
SPEAKER_01: You know, HRI, obviously, it's one of those terms.
7:58
SPEAKER_01: There's a lot there seem to be a lot of these terms in and around robotics.
8:01
SPEAKER_01: Humanoids is another good example where everyone kind of has their own specific definition of what it means.
8:07
SPEAKER_01: But when I think about underwater robots, when I think about agriculture robots, to a certain extent, um, those things are successful when they're kind of navigating or working away from humans.
8:20
SPEAKER_00: I mean, they are, but look, this is an argument that not an argument, but a discussion I've had many times with with some people, particularly uh DARPA.
8:29
SPEAKER_00: Um when you talk about robots, you don't really want autonomous robots.
8:36
SPEAKER_00: You want super capable robots, right?
8:38
SPEAKER_00: We want robots that we can task, that they go off and they do their thing and they come back.
8:42
SPEAKER_00: So the testing is part of that.
8:44
SPEAKER_00: But I also think that that also comes into the human-robot interaction.
8:47
SPEAKER_00: So if we have the robots for farming, do we want to tell the robots that, you know, I want you to focus on picking these crops today or checking out these other crops?
8:56
SPEAKER_00: Because it can't, it still needs to somehow make some decision.
9:00
SPEAKER_00: I mean, yes, we could create the most capable, biggest AI system, but but on the other hand, that's more to break at that point.
9:08
SPEAKER_00: In some sense, it's still going to need to work with the person at some level, whether it's the instructions coming from the person, or if it has trouble, if it's having a failure, how does it explain why it's having trouble doing what it's doing?
9:19
SPEAKER_00: If it gets stuck, how does it call in for help?
9:22
SPEAKER_01: That seems to be a big potential rule for large language models, right?
9:26
SPEAKER_01: Is increasing natural language interaction between robots and people.
9:30
SPEAKER_00: We've done a little bit of exploration in that space.
9:32
SPEAKER_00: So in my lab, we've been looking at ways to have robots explain failures that they're making or that they're about to make.
9:39
SPEAKER_00: So we've used behavior trees and then assumption checkers that came out of Brigham Young University to look at whether if we're already violating assumptions to begin with, then we could say, okay, we're about to fail.
9:49
SPEAKER_00: So if we're going to do those sorts of things, then we could have templated answers, right?
9:55
SPEAKER_00: Templated answers sometimes sound very stilted, poor English, and then people say, well, we don't really trust them because it doesn't sound like English.
10:01
SPEAKER_00: So we started playing with LLMs.
10:03
SPEAKER_00: And and so instead of templated language, we tried using LLMs to generate the language.
10:07
SPEAKER_00: And what we found in our first study is actually people didn't understand the explanations cutting out of the LLMs.
10:12
SPEAKER_00: And I think the issue there was not so much that the LLM wasn't good.
10:16
SPEAKER_00: Um, I think we prompted it incorrectly because we had given it what the robot could do.
10:21
SPEAKER_00: We had passed the behaviors, we had passed, you know, we basically gave it a developer's knowledge and we asked it to explain to a person who wasn't a developer.
10:30
SPEAKER_00: So they couldn't understand as well.
10:32
SPEAKER_00: So then we we tried another experiment where we actually made the prompt saying, you know, explain this to more of a novice.
10:38
SPEAKER_00: And we had a third condition where we pass in a picture of what the environment around the robot looked like, and then it could ground its explanations, and then people understood them better.
10:46
SPEAKER_00: So I think there is space in LLMs.
10:49
SPEAKER_00: I I think my concern with LLMs, of course, is how do we know that it's correct?
10:54
unknown: Right?
10:54
SPEAKER_00: Because if you're going to have the system explain something to a robot, you want to uh to a person rather about a robot, you you want it to be explaining it correctly and not making things up.
11:05
SPEAKER_00: So so we need to be pretty careful about that.
11:07
SPEAKER_01: Trust is a really interesting word in this context, and it seems like a lot of the efforts to make robots more trustworthy or or to allow people to trust robots more, or to make them more human-like.
11:20
SPEAKER_01: So in this case, right, it's it's um it's speech, it's it's natural language conversation.
11:24
SPEAKER_01: Uh, in other cases, I mean, and you made a point off the bat to say that you know you you're not really operating in social robots, but is there a way to make systems more trustworthy that isn't necessarily just turning them into a humanoid robot?
11:39
SPEAKER_00: Yeah, and I mean the reason I'm not working on social robots is I feel like there's so much to do just on the capabilities of robots.
11:47
SPEAKER_00: So if we're if I I don't feel like robots are fully where they where they need to be before we can really have them be out and be social entities.
11:55
SPEAKER_00: Um and it in terms of trust, I mean I don't think you have to be social to be trusted.
12:00
SPEAKER_00: One of the things I've been thinking about lately is you know, we've done a lot of work.
12:03
SPEAKER_00: My my lab's been working on trust for you know the last 10, 15 years, and we've looked at whether people trust the robot system or not, and and and you know, what happens in failures and what happens to the trust in failures.
12:16
SPEAKER_00: But I but one my PhD student, Greg LaMaser, has done an experiment recently, and what we found is that the robot can fail a lot, and the people can still trust it.
12:29
SPEAKER_00: Because if they understand that it's going to fail, they can trust it.
12:34
SPEAKER_00: I mean, if if if you tell me this robot is gonna do task A 100% of the time and task B 0%, I can trust it to do task A and never allow it to do task B.
12:43
SPEAKER_00: So we're starting to think now about in addition to trust, what does it mean?
12:49
SPEAKER_00: How do you how do you measure or quantify capability of the system?
12:55
SPEAKER_00: Because I think I think there's going to be a capability component.
12:58
SPEAKER_00: And how do you how do you tell people what the capability of a robot is?
13:02
SPEAKER_00: Could we prime users to know what the capabilities of the robots are?
13:04
SPEAKER_00: Can you do that quickly?
13:05
SPEAKER_00: So when they come up to you as a robot, they're pretty well calibrated as what the robot could do right away.
13:11
SPEAKER_01: I'm thinking of like C3PO all of a sudden and him always giving like the odds of of everything.
13:16
SPEAKER_01: That seems like the most I I guess the most obvious way on the face of it is just to tell people, you know, roughly what percentage of the time you succeed or fail in something.
13:26
SPEAKER_00: You you can, but you know, most people, let me just say most Americans are not really good with probabilities.
13:31
SPEAKER_00: Sure.
13:32
SPEAKER_00: Understand them, right?
13:33
SPEAKER_00: I mean, there's a large probability that Americans are going to die of a heart attack, but some are afraid to get on a plane flight because I think that's how they're going to die.
13:39
SPEAKER_00: But the probability of that is quite low.
13:41
SPEAKER_01: I should say that nobody trusted C3PO more when he tossed out those probability numbers.
13:46
SPEAKER_01: So you might be onto something.
13:47
SPEAKER_00: Right.
13:47
SPEAKER_00: So I don't know the problem, but but the interesting thing is in Greg's study that he's done recently is we found that people can pretty much identify those probabilities.
13:55
SPEAKER_00: They're not spitting them out as numbers, but we can see it coming in in the data that when we go from 100% to 70% to 50 to 30 to zero, it the bars are right there.
14:05
SPEAKER_00: People are recognizing how often it works.
14:07
SPEAKER_00: So I think people are are seeing it, but I don't necessarily think they're modeling it as probabilities.
14:12
SPEAKER_00: Um, so the so the question is, how do we get people to understand what the robots can do and when they can do it?
14:17
SPEAKER_00: I'm not sure, you know, for us, probabilistic is easy, but but for most Americans, I don't I don't know that it is.
14:23
SPEAKER_00: Um outside of the US, they're probably better at math.
14:26
SPEAKER_01: Like using the Roomba example, right?
14:28
SPEAKER_01: That's something that there's an expectation when you spend money on a robotic vacuum that there are certain things that's that it's not going to be able to do at all.
14:37
SPEAKER_01: And there are certain things that it's probably not going to be able to do particularly well.
14:41
SPEAKER_01: But as long as it does the majority of the things that you wanted to do a majority of the time, then most people will consider it reliable.
14:49
SPEAKER_00: Yeah.
14:50
SPEAKER_00: I mean, when I open that bin and there's a lot of stuff in the bin, I feel like it did its thing.
14:55
SPEAKER_01: Yeah.
14:55
SPEAKER_00: Before that, the scuba, which sadly they don't sell anymore, that water would come out really dirty.
15:00
SPEAKER_00: It was amazing.
15:00
SPEAKER_00: It was really good.
15:01
SPEAKER_00: So I think that's probably why they don't sell it anymore.
15:04
SPEAKER_00: But it was like, wow, really clean my floor.
15:06
SPEAKER_00: Look at this water I'm emptying.
15:08
SPEAKER_00: You know, and the roombas that I continue to buy are the the less smart roombas.
15:13
SPEAKER_00: I don't care if it's mapped to my environment.
15:14
SPEAKER_00: Um I'm happy with it bouncing around.
15:16
SPEAKER_00: That's that's enough to get my floor clean.
15:18
SPEAKER_00: But I don't know, it'd be interesting, and I'm sure there have been studies done on whether people I'm I'm I'm sure they must like the wrapping.
15:24
SPEAKER_00: Otherwise, I why would i robot have done it, right?
15:27
SPEAKER_01: That's interesting in and of itself, in that your expectation is just to get the floor clean.
15:33
SPEAKER_01: And if it's doing so in like a random motion, then that's totally fine as long as it does what it's supposed to do in a given time.
15:40
SPEAKER_00: Yeah, I mean, honestly, most of the time I don't care how long it takes, I set it off and I leave.
15:43
SPEAKER_01: Do people necessarily want complexity?
15:45
SPEAKER_01: Do they they over expect complexity from robots?
15:49
SPEAKER_00: I think the successes that we've had in robots is where we can narrow things down, right?
15:52
SPEAKER_00: And we can make them work in very particular situations.
15:55
SPEAKER_00: Um, you know, certainly the Roomba was incredibly successful, right?
15:59
SPEAKER_00: Tens of millions of robots were sold.
16:01
SPEAKER_00: You know, that said, Waymo's out there with cars, and I've had fun riding in them in San Francisco and Phoenix, and you can call them.
16:09
SPEAKER_00: Uh and so more and more robots are out there, but but still I think they are they're limited, right?
16:15
SPEAKER_00: I mean, it it's very clear when you ride the cars in San Francisco that they've got that model down to pretty much millimeters, right?
16:21
SPEAKER_00: You you can see it when you're when you're watching the screen when they're driving.
16:24
SPEAKER_00: So it's it's not going off into uncharted territory to to be driving.
16:29
SPEAKER_00: That's that's how we deploy robots, right?
16:30
SPEAKER_00: You have to create some sort of limitations.
16:33
SPEAKER_01: I think though there is people perhaps trust something less when there is this.
16:39
SPEAKER_01: I mean, it's it's almost counterintuitive, but the fact that somebody like is sitting in the driver's seat and can take over almost leads people to trust it less because they know that that's somehow implicit that it's not getting it right all of the time.
16:51
SPEAKER_00: I mean, rides I took, there's nobody in the driver's seat.
16:54
SPEAKER_00: I mean, I believe that there was one time the robot was clearly taken over.
16:59
SPEAKER_00: Um we sat for about 60 seconds and then we moved forward.
17:01
SPEAKER_00: So I think it was taken over remotely.
17:03
SPEAKER_00: I don't know for sure.
17:04
SPEAKER_00: But but I think I mean it was actually really interesting because it was a group of roboticists and we were using them and and some people were still uncomfortable with it.
17:13
SPEAKER_00: And my feeling was look, if I'm not willing to get into a car and let it drive me, a Waymo car.
17:20
SPEAKER_00: Totally different for other brands.
17:22
SPEAKER_00: Yeah.
17:24
SPEAKER_00: I mean, I want to have trust in the company that's putting together and believe that they've actually tested it.
17:29
SPEAKER_01: That's fair.
17:29
SPEAKER_01: That's fair.
17:30
SPEAKER_01: And and you and you do have a history with with Google.
17:32
SPEAKER_01: You know, you know, again, I was watching this teleop thing that you're doing, and that was like what, 15 years ago that you were testing out these uh uh teleconference robots over at Mountain View.
17:42
SPEAKER_00: Yep, yep.
17:43
SPEAKER_00: That was a couple of my students worked there, and we tested different ones in Mountain View.
17:47
SPEAKER_00: Um, gosh, that was a really long time ago.
17:50
SPEAKER_00: Um, and that actually led into Kate Sue's PhD thesis, where she had taken a Vigo robot and made it a lot more autonomous.
17:56
SPEAKER_00: And that was with the Hawaiian shirt on the robot.
17:58
SPEAKER_00: What's been really fantastic about my career in robotics is I've had lots of opportunities to deploy robots in lots of different places.
18:06
SPEAKER_00: My students competed in a competition at the Johnson Space Center on the on the rockyard one year, and that was fantastic.
18:13
SPEAKER_00: We got to go visit the moon rock facility with people who had been working with the rocks since they had been brought back from the moon.
18:18
SPEAKER_00: And so it's a fantastic room.
18:21
SPEAKER_00: I always say I have the best toys, and then I've got this robot playground to go put them in.
18:26
SPEAKER_01: Yeah, I I I know and I know a lot of people are hesitant to give predictions, but I'm I'm wondering if if you and I had spoken about the same subject in 2010, about the prevalence of teleconferencing robots in 2025.
18:39
SPEAKER_01: Would you expect there to have been a lot more out in the world than there are now?
18:43
SPEAKER_00: I think COVID was an inflection point, right?
18:45
SPEAKER_00: Actually, I think in 2010 I would have expected there to be more.
18:48
SPEAKER_00: And after COVID, what we saw is that it just didn't really find find its niche.
18:53
SPEAKER_01: Yeah, it wasn't as much of an inflection point as it could have been, really.
18:57
SPEAKER_00: I mean, it in some sense I think it w it went the other direction, right?
18:59
SPEAKER_00: We all went into Zoom.
19:00
SPEAKER_00: There was no point to have a robot in a place when nobody was in those places.
19:05
SPEAKER_00: Um, so so that was uh I think kind of a strong indicator against telepresence robots, which was kind of unfortunate because it felt like the the the field was moving along and and with COVID, it just there wasn't a use case for them.
19:19
SPEAKER_01: Yeah, is there a reason, you know, like again, it it it seems like that technology has been around for a really long time.
19:25
SPEAKER_01: Like I again, I remember visiting iRobots office in Bedford a long, long time ago, and they had some version of that too.
19:32
SPEAKER_01: Why haven't they really taken off in a meaningful way?
19:35
SPEAKER_00: Well, I mean, I think if you go back 20 years, there wasn't good enough internet uh in most buildings to support it, right?
19:42
SPEAKER_00: So I mean maybe 30 years ago.
19:44
SPEAKER_00: So I think that was one of the problems.
19:45
SPEAKER_00: You just couldn't get the video to go.
19:47
SPEAKER_00: I I think beyond that, I mean, there's an expense, right?
19:52
SPEAKER_00: And there's this notion of there's also it it feels a bit um uneven.
20:01
SPEAKER_00: So if if I'm in a space and there's a telepresence robot, somebody can come into my space.
20:05
SPEAKER_00: And that was actually at Willow Garage, they had developed the telepresence robot, and they had a rule that if you were in our space, you had your camera on.
20:13
SPEAKER_00: Right.
20:13
SPEAKER_00: So they they required people who were driving in their space to have cameras on because they didn't feel like it was fair that people could come in and see them, but they couldn't see who was driving the robot.
20:21
SPEAKER_00: So so it's kind of this one-sided.
20:23
SPEAKER_00: I think when you start looking at people who are older, um, would they want to have a telepresence robot?
20:28
SPEAKER_00: There's sometimes they're a little bit leery about well, can my kids just come into my space whenever they want?
20:34
SPEAKER_00: Yeah.
20:34
SPEAKER_00: Whereas with the phone, you can answer or not answer.
20:37
SPEAKER_00: And I you could, of course, you can mute a telepresence robot.
20:40
SPEAKER_00: I think that, you know, some of the use cases for telepresence robots where I think it could be really valuable is, you know, Kate Suey, when she was doing her doctorate, was looking at possibilities for allowing people with disabilities to travel to museums through telepresence robots.
20:55
SPEAKER_00: So if they couldn't get there, they could be there.
20:57
SPEAKER_00: And she was looking at how you have interactive exhibits that the robots could interact with, so somebody could actually be interacting with them remotely.
21:04
SPEAKER_00: So I think there are, I think there are good use cases.
21:06
SPEAKER_00: I know Vigo back in the day was working with letting students go to school using them.
21:11
SPEAKER_00: So if students were immunocompromised, they could use them.
21:13
SPEAKER_00: So I think there are cases for them, but I think we got so used to using Zoom and other video conferencing systems that, you know, do we need to have a presence somewhere else?
21:27
SPEAKER_01: Yeah, and you you touched on something, again, getting back to that conversation or that uh talk that you were giving around telepresence.
21:34
SPEAKER_01: That I mean, it once it's one of these things that once you say it, it's obvious on the face of it.
21:39
SPEAKER_01: But when we think of robots, telepresence robots, we tend to think of the experience that the person is having who's remotely operating it, but not so much the people in those meetings who necessarily have to sit with the meetings, sit with the robots in the meetings.
21:52
SPEAKER_01: That like that isn't necessarily as good of an experience as the other way around.
21:56
SPEAKER_00: Right.
21:56
SPEAKER_00: And I and I think that the the systems that we had tested in Google.
22:00
SPEAKER_00: Years ago, neither of them had any mapping capabilities.
22:03
SPEAKER_00: And you'd mentioned the one that iRobot had, Ava, it spun out into its own company, Ava Robotics.
22:08
SPEAKER_00: That one, you know, has a lot of mapping.
22:10
SPEAKER_00: So that does have the capability that once you have that meeting and you're done with that meeting, the robot can go to the next location it needs to go to.
22:19
SPEAKER_00: Whereas if you're the remote person using it and I have to drive it to the next location or drive it back to the dock, when a meeting ends, I can't just get on my email or go to my next meeting.
22:29
SPEAKER_00: I have to take care of the robot.
22:31
SPEAKER_00: So there's issues even for the remote person.
22:34
SPEAKER_00: Um, you know, people did feel like they were more present in the robot, that people paid attention to them more in a meeting.
22:40
SPEAKER_00: Um, especially it worked out really well when there was like one person remote and then there was a big group because it kind of brought them into the group.
22:47
SPEAKER_00: When there were two larger groups, then it it didn't quite have the same outcome.
22:51
SPEAKER_01: So how long were you at Willow Garage for?
22:54
SPEAKER_00: I was not at Willow.
22:54
SPEAKER_01: I was just You were not at Willow Garage.
22:56
SPEAKER_01: Oh, okay, okay.
22:57
SPEAKER_00: I was just that's an example of them doing that work at Willow, but I was not there.
23:00
SPEAKER_01: Yeah.
23:00
SPEAKER_01: I mean, you've been on on really on the research side um for your entire career.
23:05
SPEAKER_00: Yeah, yeah.
23:06
SPEAKER_00: I graduated from from my PhD, uh spent a year at Boston College, um, and then moved up here to UMS Low and I've been here 24 years and now I'm making a big move to another university.
23:16
SPEAKER_00: Um so so so yeah, uh unlike my my advisor who's done many, many companies, um, I've not.
23:24
SPEAKER_01: Yeah, yeah.
23:25
SPEAKER_01: Who was uh we had on the show is Rod Rod Brooks.
23:29
SPEAKER_01: Um he's uh you know, say stayed in academia for a little while.
23:33
SPEAKER_01: And I'm sure that there have been opportunities.
23:35
SPEAKER_01: I'm sure that you've had a lot of students come through with, you know, uh uh technologies that could be commercialized.
23:42
SPEAKER_01: Um, was that ever a move that you were interested in making, even if just maybe dipping a toe in?
23:49
SPEAKER_00: For me, the exciting part about academia is the ability to do lots of different things, um, as opposed to you know really putting everything down on one thing.
24:00
SPEAKER_00: And and I mean, I've certainly had comp students go and work in a lot of different companies.
24:04
SPEAKER_00: Um I so I think stuff in the lab does spin out at least the ideas too.
24:08
SPEAKER_00: But there hasn't been something where I've been like, yes, I have to go and start a company with this.
24:13
SPEAKER_00: It it just I I haven't found it yet.
24:16
SPEAKER_00: Maybe maybe the thing will be there.
24:18
SPEAKER_00: But I I really like the the freedom to try all the things.
24:23
SPEAKER_00: And you know, back in grad school, um, I worked at the digital uh research lab in Cambridge for a few summers.
24:30
SPEAKER_00: And it just what you need to do to show that your research is of value to the company, that things can be canceled at any time.
24:40
SPEAKER_00: And, you know, yeah, I have to find funding to pay for research, but but there are ways.
24:45
SPEAKER_00: I mean, a student can be a TA and still be doing the research that you can't find funding for, right?
24:48
SPEAKER_00: So you can still do the ideas that you can't get money behind them if you think they're good good ideas and want to pursue them.
24:53
SPEAKER_01: What are those pressures like?
24:55
SPEAKER_01: I mean, you know, again, I'm kind of coming at this from the more corporate side of things, but as far as you know, you you are a student having a good idea, is is it something where you necessarily like need to reach out to the people giving the grants or giving the investments to get it past them?
25:10
SPEAKER_01: Or are you really able to just sort of like work on a project and see it to fruition based on whatever grants you have?
25:16
SPEAKER_00: Yeah, I mean, I think it depends on on where you stand in the funding cycles.
25:20
SPEAKER_00: I mean, certainly, you know, we'll we'll write proposals for things that we've got ideas for.
25:24
SPEAKER_00: Uh I'll bring in students for things I have ideas for.
25:26
SPEAKER_00: But there's always, you know, I had an another student, uh, Mark Massiri, and he was super interested in doing multi-touch to control robots, and we could not find a funder who wanted to pay for it.
25:38
SPEAKER_00: Um, so we found other ways to to cover his time and and and and yeah, he did a great thesis, uh, went off and worked for NASA and Docker for a while and now is in Australia.
25:47
SPEAKER_00: Uh and and so I think that you know there are there are ways.
25:52
SPEAKER_00: I mean, I think I think the thing that we become really good at as professors is telling a story about how everything in our lab fits together, even though the students don't think it is fitting together at all.
26:00
SPEAKER_00: But I mean I think the funding funding may look different moving forward.
26:06
SPEAKER_00: We'll see.
26:07
SPEAKER_00: Hopefully.
26:08
SPEAKER_00: Hopefully there'll still be US funding for for robots and AI.
26:12
SPEAKER_00: I think that there's certainly indications that the administration is is supportive of AI, which which is good.
26:18
SPEAKER_01: Yeah.
26:18
SPEAKER_01: Is there, I mean, is there generally more wiggle room if it is like a government grant versus a corporation coming in?
26:25
SPEAKER_00: It really depends.
26:26
SPEAKER_00: Some corporations you know will give kind of unrestricted gifts, and then that that's a wiggle room as well.
26:31
SPEAKER_00: So I mean, I mean it's it's really finding a way to to enable.
26:36
SPEAKER_00: I mean, I really look at my role as enabling my students to do the things they want to do, right?
26:40
SPEAKER_00: I need the resources for them to help them out.
26:43
SPEAKER_01: Every time I'm writing something up about a research project, and then you know, finally come to the end, and then you start talking about like, oh, these are the applications that it could have in the future, you you you you you hook on to it first because it looks like a new a uh cool technology, and then you get to the actual like efficacy and the usefulness of the product at the end.
27:02
SPEAKER_01: Is that is that something, is that kind of real-world usage, whether it's commercialization or anything else, something that's at play from the beginning of the project?
27:13
SPEAKER_00: I mean, it should be.
27:16
SPEAKER_00: If you're designing things that are going to work with people, uh you should certainly be talking to your intended user groups from the beginning and then working with them.
27:23
SPEAKER_00: So we've certainly done a lot of work with you know participatory design and trying to get feedback from the intended groups.
27:29
SPEAKER_00: Of course, we all in academia fall back on our convenience populations of testing things with undergraduates because that's that's what we have.
27:36
SPEAKER_00: Um, but but really, if you want to do good work and think about what could be deployed, you need to be working with the intended user groups.
27:43
SPEAKER_00: I think that one difficulty in my world, I do a lot of work in assistive technology, and it's just quite difficult in the US to get insurance to pay for a lot of these systems.
27:54
SPEAKER_00: Um and so, you know, people have been working on robotic wheelchairs.
27:57
SPEAKER_00: I mean, I did it for my thesis, you know, whatever it was a long time ago, uh, over 25 years ago, and and and people have been doing it since, but there's no real push to to get this out there, right?
28:10
SPEAKER_00: So if you need assistance with driving and your power chair, then you probably need assistance and you need a person to help you, and then they just say, well, why do you need a powered wheelchair if you were paying for someone to help you?
28:20
SPEAKER_00: So there's sometimes it's hard to find the the business model for some of these technologies as well.
28:27
SPEAKER_01: Yeah, yeah.
28:28
SPEAKER_01: I I feel like every couple of years I see like a cool new robotic wheelchair, and you know, it doesn't necessarily go anywhere.
28:36
SPEAKER_01: Um it seems like exoskeletons, though, like at least to a certain extent, certainly um in commercial settings are are finding their place now.
28:45
SPEAKER_00: We've been doing a lot of quick looks for the army on different types of systems.
28:49
SPEAKER_00: So there certainly are a lot of systems being developed.
28:52
SPEAKER_00: Um so so we do we do a lot of, you know, we bring in just you know six or eight subjects to to try them out in different cases.
28:59
SPEAKER_00: Because of course, all the the companies say, oh, it could do X, Y, and Z, but we're trying to do a consistent evaluation across all of them so the Army can be taking a look at them.
29:08
SPEAKER_00: I think that even then, you know, the there's certainly the military has been working on trying to get exoskeletons now for what decades, right?
29:17
SPEAKER_00: Yeah.
29:18
SPEAKER_00: Um I think that, you know, one of the maybe the difference now is that they're much more focused on one or two joints.
29:26
SPEAKER_00: Uh you don't see sort of the whole body sarcos big exoskeleton that that you saw 25, 30 years ago.
29:32
SPEAKER_00: And even then, you know, there's there's a lot of work trying to do predictive modeling about what the person is doing so you can provide the right assistance.
29:41
SPEAKER_00: But that's hard, right?
29:43
SPEAKER_00: If you if you think of having an exoskeleton where you're running and then you stop, even the best prediction in the world, there's there's just a lag, right?
29:51
SPEAKER_00: So, you know, you might have to make sure you relax your knees so you don't get that last push after you stop.
29:57
SPEAKER_00: So it these are these are really hard problems to solve.
30:01
SPEAKER_01: My father was diagnosed with Parkinson's last year, and I was doing some research on some of the exoskeletons around Parkinson's, and then like that presents its own challenge and is that there's a way that people move forward and there's kind of a special sort of predictive algorithm that you need in order to continue that that motion going forward.
30:19
SPEAKER_00: We don't think anything about how we're moving and we stop and we turn, and but you know, the system has to get that through our muscle motion, right?
30:28
SPEAKER_00: And and it it has to have that second or two to to get the sensor readings and process it.
30:34
SPEAKER_00: So so it is a really hard problem.
30:37
SPEAKER_00: I think that you know, people are making some strides in that, but but but there's a lot to still be done there.
30:43
SPEAKER_00: I mean, I think it's I think assistive robots is a really interesting thing to be working on.
30:48
SPEAKER_00: I think that you know, from an HRI perspective, you know, some robots are far away from you.
30:52
SPEAKER_00: Um some of them you ride on, and then some of them you're you're really integrated with them, like you are with the exoskeletons that creates a very different sort of human-robot interaction.
31:01
SPEAKER_01: As you're making this move to to Amherst, um will there be I I guess you kind of have to just start over with a new group of students.
31:08
SPEAKER_01: Is that right?
31:09
SPEAKER_01: Or do you get to take some folks with you?
31:12
SPEAKER_00: Um so I've been chair of of the minor school of computer information sciences here.
31:18
SPEAKER_00: So I actually didn't pick up new students for the last three years, so I'm finishing up my students here.
31:22
SPEAKER_00: Um so it really is kind of a fresh start over there.
31:25
SPEAKER_00: I'll be finishing my students, they'll be staying here.
31:27
SPEAKER_00: But one of the things I'm going there is to help them grow their program.
31:30
SPEAKER_00: They have, they've got several faculty doing good work in robotics, um, but they want me to come in and you know help expand, you know, bring the nerve center there to think about how we could have more academic programs and robotics.
31:43
SPEAKER_00: And and I think it's it's interesting to look at bringing robotics, you know, Pioneer Valley.
31:49
SPEAKER_00: So Massachusetts, you know, you've got the you've got Boston and Cambridge, and then you've got 1.8 around it.
31:54
SPEAKER_00: Um and that's mostly where most people want to stay.
31:56
SPEAKER_00: Then you have 495 and Lowell's out on 495, and then you go farther west, and then you're in Amherst.
32:02
SPEAKER_00: More people want to be near Boston, but I think that there are a lot of people that, you know, we could bring companies out out west, and there's there's certainly people working on that.
32:13
SPEAKER_00: And I think it'd be interesting to see how we can, you know, expand robotics out there in the five college area.
32:20
SPEAKER_01: Obviously, the conversation around brain drain has changed quite a bit post-pandemic, especially with people, you know, increasingly working remotely.
32:27
SPEAKER_01: But what what is what is your pitch?
32:28
SPEAKER_01: What's your pitch to get people to sort of move outside of the greater Boston area?
32:32
SPEAKER_00: So I'm getting an apartment in Amherst.
32:34
SPEAKER_00: So I actually am not moving outside the greater Boston area, at least not the first year.
32:38
SPEAKER_00: I think there's a lot of talent.
32:39
SPEAKER_00: And I think that, yeah, real estate's cheaper, right?
32:44
SPEAKER_00: If you want to have a company out there, I think that there are five great colleges in the area out there.
32:52
SPEAKER_00: I think that's an opportunity for growth.
32:53
SPEAKER_00: And I think they're looking, you know, Massachusetts is doing a lot of investment in AI and robotics, and this the Commonwealth is interested in investing outside of Boston as well, right?
33:03
SPEAKER_00: They certainly invest in in Boston, but but they're looking to have all of the areas in Massachusetts uh benefit economically.
33:10
SPEAKER_00: So I think that I think there are a lot of people who would be interested in working in robotic companies out there, uh, good quality of life, you know, bike trails, hiking.
33:23
SPEAKER_00: Cheaper.
33:24
SPEAKER_00: I don't say I'm not sure the real estate is that much cheaper around the university from what I've looked at.
33:29
SPEAKER_00: Um but talk to me in a year and I'll give you a better, a better sales pitch on Amherst once I've been out there for a little bit longer.
33:36
SPEAKER_01: It's gonna be an interesting ramp-up period too, right?
33:38
SPEAKER_01: I mean, if there are if there isn't really an existing or you know, as much of an existing robotics um program, it's gonna take a few years, really, I guess, to get to a critical mass for you.
33:50
SPEAKER_00: Well, yeah, I mean, look, there are six core strong faculty and and then many others working in robotics.
33:56
SPEAKER_00: Uh when I started at Lowell, I was the only person doing work in robotics.
34:00
SPEAKER_00: So, you know, it took a little bit of time.
34:02
SPEAKER_00: I I think the biggest thing for me is they're renovating a lab for me.
34:10
SPEAKER_00: Uh don't get me wrong, it's nice to have a renovated lab, but I won't be ready for six to twelve months.
34:14
SPEAKER_00: And so I guess it's it's putting me back to the beginning of like, holy smokes, I'm gonna have a new lab.
34:21
SPEAKER_00: And but but maybe you know that that's exciting because it gives you a chance to rethink some things and what are the things you want to keep and what are the things you want to change and what do you want to be focusing on.
34:32
SPEAKER_00: Um, I'm really excited about working with my new colleagues out there.
34:34
SPEAKER_00: They also have, you know, obviously very strong groups in machine learning, uh, the Turing Award, just went to antibardo uh for reinforcement learning, and and and many people doing AI and computer vision.
34:46
SPEAKER_00: So I so I I think there's you know a strong group of researchers to work with out there.
34:51
SPEAKER_00: And then they have you know nursing and physical therapy and uh and agriculture, which we don't have here at Lowell.
34:58
SPEAKER_00: So I so I think it's and I'm not abandoning my colleagues here.
35:03
SPEAKER_00: I'm planning to still work with the people I've been working with here, right?
35:06
SPEAKER_00: Yeah.
35:06
SPEAKER_00: So so I it's it's a way to expand what I'm doing.
35:10
SPEAKER_00: Um, but it's I I guess I want I want to be challenged.
35:14
SPEAKER_00: I want to I want to grow some new stuff, um, you know, creating an undergrad program.
35:19
SPEAKER_00: Thinking one of the things I've been thinking a lot about is here in Massachusetts, now community college is free for everybody.
35:24
SPEAKER_00: And how do we create a robotics program at the undergrad level that takes somebody who went to community college for two years, maybe in a robotics technician program that doesn't have a lot of math?
35:34
SPEAKER_00: And and how do we think about turning that into a four-year degree?
35:38
SPEAKER_00: And it's it's not simple, but but I think it'd be an interesting problem to look at.
35:43
SPEAKER_01: Yeah, well, and that that's really the start of a larger, really important conversation around like upskilling, right?
35:49
SPEAKER_01: Or around as more robots are entering the workforce trying to find some of these ways to get people who don't necessarily have a background in programming or math robotics to learn how to work alongside these systems.
36:02
SPEAKER_00: But I think even working alongside them, is there a way to have people, you know, is there a better way to program the robots so that people who don't have programming can in some sense, you know, certainly there are teach pedants and those sorts of things.
36:16
SPEAKER_00: Um, but uh but are there other ways?
36:18
SPEAKER_00: Are there ways that people could give language instructions to the robots so that they could, you know, kind of prompt an LLM in some sense, right?
36:27
SPEAKER_00: Um, but of course you'd have to make sure that things work.
36:30
SPEAKER_01: Yeah, I mean you must have learned a lot on that front when you're doing the the art botics program, right?
36:34
SPEAKER_01: As far as how to you know appeal to sort of a larger group of people and people who aren't don't necessarily have that robotics background.
36:42
SPEAKER_00: Yeah, I mean, I think actually artbotics goes back uh that program we were working with UMass Amherst, so it's at least ties from.
36:49
SPEAKER_00: Um, but but I you know, I think that, you know, the one of the things that we thought a lot about in artbotics was that students come to college, like, I'm gonna be in computer science, I'm gonna do video games and robots, and we set them down in computing one and make them do compound interest tables.
37:06
SPEAKER_00: Um and and so instead in our botics teach people at four loops.
37:10
SPEAKER_00: We had a small round chassis and we had students basically, you know, program a pattern and then loop it so you can create spirograph-like patterns.
37:19
SPEAKER_00: So so there was there was a reason to do loops because you wanted to see something happening multiple times um as opposed to compound interest, which especially when interest rates were really low.
37:29
SPEAKER_00: So so I think that you know we our botics also really focused on kind of sensors and thinking about how you wanted it to interact with people and and the idea that also sensors help you not to have the motors wear out because they're not running all the time.
37:44
SPEAKER_00: So so I think that there are ways to go towards that.
37:48
SPEAKER_00: I think that we as a robotics field, you know, certainly with you know probabilistic robotics, have have moved super math, which is important.
37:56
SPEAKER_00: I mean, these these are important things that we're doing, but are there other areas of it that we could open up?
38:01
SPEAKER_00: I've I've long wanted to think about how we could have you know robotics classes for MBAs.
38:08
SPEAKER_00: Because those are the people who are going to be buying the robot systems.
38:10
SPEAKER_00: What do they need to know about them?
38:11
SPEAKER_00: What do they need to understand about them?
38:12
SPEAKER_00: What questions should they be asking?
38:14
SPEAKER_01: Um Yeah, or the people, you know, becoming like the CTOs of robotics companies as well, in a lot of cases, like executives coming over.
38:22
SPEAKER_01: So what is that?
38:23
SPEAKER_01: I mean, you know, there's obviously there is some math background for a lot of MBAs, but what how do you bridge that gap?
38:29
SPEAKER_00: Yeah, I mean, I look, I I don't think everybody, you know, needs to be programming them, not everybody needs to be, you know, doing linear algebra, but I but I think that you know, even just knowing what what questions to ask and and and to think about, you know, maybe maybe uh maybe I'll teach a whole bunch of MBAs about how to what can you do to make this robot fail during the demo?
38:51
SPEAKER_00: Um maybe that would be a great outcome because I go to companies that they break things.
38:56
SPEAKER_00: You can't just accept.
38:57
SPEAKER_00: And I think this is kind of goes back to what the army is doing with the quick looks is you can't accept what companies are telling you, right?
39:03
SPEAKER_00: You can't, you know, or or if you look at like small unmanned aerial systems, then they might say, oh, it flies for 15 minutes, but it might fly 15 minutes when it's not running any of its sensors, right?
39:15
SPEAKER_00: So you need to be able to test those against one another and kind of have a standard way of looking across all of these claims.
39:22
SPEAKER_00: Um so so maybe that's the thing for MBA.
39:25
SPEAKER_00: So this is these are things I'm gonna start thinking about.
39:28
SPEAKER_00: Um I'm going to visit University of Michigan in September because they've they've created a good undergrad program.
39:34
SPEAKER_00: So I'm gonna go visit them and look at that.
39:36
SPEAKER_00: Um and thinking about how we're gonna do that.
39:39
SPEAKER_00: And and what is it, what does it look like to create a robotics master's program when visas are tough right now?
39:48
SPEAKER_01: Yeah.
39:49
SPEAKER_00: So so so how what what does that look like?
39:52
SPEAKER_00: You know, will it will it change, hopefully.
39:54
SPEAKER_00: Um obviously we need to bring more talent into the US.
39:58
SPEAKER_00: So but but you know, even how do we focus on community colleges?
40:02
SPEAKER_00: And then that's focusing on on more US students too, as well.
40:05
SPEAKER_00: So I so I think everybody, everybody needs to be in robotics, of course, um, and finding the pathways for everybody to get there.
40:12
SPEAKER_01: I always hear these great stories, especially from um folks who are in ag tech robotics, about really sort of like getting out, like literally getting out into the field and meeting farmers and um, you know, just sort of the the the new dimensions that you learn about the the work that you're doing.
40:30
SPEAKER_01: Is that uh as you're taking a look into agriculture, is that sort of part of your own plan for yourself is to kind of get out of the laboratory a little bit more and actually like interact with these more in the field?
40:41
SPEAKER_00: Yeah, I mean, I I think that that goes back to what we were talking about is you know, getting talking to the intended user population, right?
40:48
SPEAKER_00: You're gonna be developing systems that you need to find out what what the problem spaces are.
40:53
SPEAKER_00: And and certainly, you know, Joe Jones, um, one of the creators of of the Roomba, he had done harvest automation and gone out and talked to people about you know moving pots in the commercial growers and stuff like that.
41:04
SPEAKER_00: So I think you know, it takes a lot of effort to go do that.
41:08
SPEAKER_00: Maybe that's a really nice crossover with those business students.
41:11
unknown: Yeah.
41:12
SPEAKER_00: Right.
41:12
SPEAKER_00: Because they're starting to identify what those problems are and those issues that need to be solved.
41:18
SPEAKER_01: Yeah, so so we are starting to come up on time, but people will always ask you, and you'll always ask yourself during different points of your career, you know, if I had to do this all over again, what would I do?
41:29
SPEAKER_01: But like you're affording yourself or the schools are affording you the opportunity to do that to really in in some way start from scratch with you know decades of knowledge.
41:38
SPEAKER_01: What are what are some of the things that in starting the second nerve center that you're not going to do over again?
41:48
SPEAKER_00: We're not going to count on one company using it so much that it pays for everything.
41:52
SPEAKER_00: That did not work out.
41:53
SPEAKER_00: And we pivoted pretty quickly and moved more towards grants and and you know, finding more customers.
41:59
SPEAKER_00: Um, at the time we created it, there was one person, one company that said, Oh, we'll be in there four days a week.
42:04
SPEAKER_00: They were not.
42:04
SPEAKER_00: Um, and my part of that was just because it took so long to start it up.
42:08
SPEAKER_00: Uh you know, by the time you get the space and you build it up.
42:11
SPEAKER_00: So I think that that that's one thing is is thinking more of Nerve supporting people who are developing robots.
42:20
SPEAKER_00: But hopefully, you know, we've worked a lot with NIST.
42:23
SPEAKER_00: Um, and I hope NIST will still exist and be able to do this.
42:26
SPEAKER_00: They've um they there's not large money there, but they've they've really been investing in thinking about how how we test robot systems and how we validate AI and robotics.
42:39
SPEAKER_00: You know, one thing that I don't have to do all over again is because I'll still be working with the Nerve Center here at UMass Lowell, is that you know I still have a strong group here that I'll be working with as we expand to have more people out there.
42:50
SPEAKER_00: So it's it is starting from scratch and it's not starting from scratch.
42:53
SPEAKER_00: Um uh I think the craziest thing is my uh irobot atrv junior, which was the robot I bought at my startup package here 24 years ago, is moving with me to Amherst because I'm still gonna be using it.
43:06
SPEAKER_00: That robot is really solid.
43:08
SPEAKER_00: Um, and it has a full-size computer chassis inside it, so we just keep updating it.
43:12
SPEAKER_00: And that platform is it's you know, people are like, you're still using the same robot after 25 years?
43:17
SPEAKER_00: Like, yeah, it's a good robot.
43:19
SPEAKER_00: So but that's why you can't make money off of academics and robotics, right?
43:23
SPEAKER_00: We just keep things going for.
43:24
SPEAKER_01: I know you're using fetch.
43:26
SPEAKER_01: Obviously, fetch isn't really a you know going concern as much anymore.
43:29
SPEAKER_01: Um uh, you know, the uh Baxter robots were big in the academic world.
43:35
SPEAKER_01: Um what what what is it?
43:38
SPEAKER_01: Unitree.
43:38
SPEAKER_01: I know there's probably actually some issues now with some of these, like you know, the back doors and maybe working along some Chinese companies.
43:44
SPEAKER_00: Yeah, we have spots.
43:45
SPEAKER_00: Um with Unitry.
43:48
SPEAKER_00: So we have a couple of spots, and I'll be I'll be getting those as well at UMass Amherst.
43:51
SPEAKER_00: That honestly, hands down, best robot I've ever gotten out of the box.
43:55
SPEAKER_00: Um, it it just works.
43:58
SPEAKER_00: Um, so so we'll we'll be Doing that.
44:01
SPEAKER_00: The UR5.
44:02
SPEAKER_00: I mean, certainly there's a lot of you know where we are right now with with um you know that's an old old robot.
44:10
SPEAKER_00: I mean it's it's you know manipulation, right?
44:12
SPEAKER_00: It's it's a good robot, it's solid.
44:13
SPEAKER_00: The universal robots, they're solid robots, but there's a lot of different manipulators to use.
44:17
SPEAKER_00: We have we have a number of them now at NERV in what we call the Arm Ada, um, which is a little bit of a joke because we started we were working with the Advanced Robotics and Manufacturing Institute, so we called it the ARM Ada.
44:27
SPEAKER_00: Um we wanted ARM Farm, but Google got there first.
44:32
SPEAKER_00: Um lots of different robots and you know, some old, some new, but I think you know, we're using stretches.
44:43
SPEAKER_00: That's yeah, I'll be using those in my classes.
44:46
SPEAKER_00: Um and the spots.
44:48
SPEAKER_00: I I I don't know beyond that.
44:50
SPEAKER_00: I don't know if there's a platform that I'm like, wow, I really need to get this new platform right now.
44:57
SPEAKER_00: Um, I think a lot of the robot companies are are looking at you know actually developing them to put right into an application and and much less so for for us academics.
45:06
SPEAKER_00: Um and and some companies don't want to sell to academics just because they don't want us publishing something that says something bad about their robot.
45:12
SPEAKER_00: Which I get it, I do.
45:14
SPEAKER_00: Um so but if you look at now versus you know whatever, 30 plus years later in grad school, when people we had to build our own frame grabbers.
45:25
SPEAKER_00: I mean, and now you can buy all sorts of different robots and you know, tons of simulations out there and tons of companies working on it, and I mean just Ross and what that enabled.
45:36
SPEAKER_00: So I I think we're I think we're at a great place.
45:39
SPEAKER_00: I mean I ride in robots when I travel.
45:44
SPEAKER_00: It's it's pretty robot cleans my house, and I ride in robots when I travel.
45:47
SPEAKER_00: So so I think it's you know, I'm I'm bullish on robots.
45:51
SPEAKER_00: I mean it'd be kind of weird if I weren't, but but I think I think there's still more to come.
45:56
SPEAKER_00: It's gonna be great.
45:58
SPEAKER_01: And and I have to ask you, every time I hear, you know, you're talking about Armada, every time I hear a really good acronym, I wonder how much of that acronym was sort of backronymed and fudged.
46:07
SPEAKER_01: And NERV is a really good one.
46:08
SPEAKER_01: Did you have to work your way backwards into NERV?
46:11
SPEAKER_00: So we actually wanted to do NERD originally.
46:14
SPEAKER_00: So we wanted to be the New England Robotics Development Center, but at the time, Microsoft, I don't know if they still do, but they had the Microsoft Nerd Center in Cambridge.
46:22
SPEAKER_00: And so we're like, oh, we probably shouldn't overlap with that.
46:25
SPEAKER_00: And and I was saying this at a meeting, and and someone who worked at the BU at the time said, Well, what about like NERVs, you do England Robotics Validation Center?
46:35
SPEAKER_00: And it's like, well, that's like, but experimentation.
46:38
SPEAKER_00: Then so it kind of came by by talking to somebody else.
46:41
SPEAKER_00: Um, so I wouldn't say it's a backronym, it was certainly helped along, but it you know, I I love it, and people are like, Oh, you're nervy.
46:49
SPEAKER_00: Uh, we do occasionally get calls about people with nerve problems, and I feel really bad that we use the name Nerve.
46:55
SPEAKER_00: Um, but but no, it it kind of went from nerd to nerve without the and nerve with the.
47:01
SPEAKER_01: Professor Yanko, thank you so much for joining us.
47:04
SPEAKER_00: Thank you very much.
47:04
SPEAKER_00: Enjoy talking to you.
47:06
SPEAKER_01: Me too.
47:08
SPEAKER_01: Thank you so much, Professor Yanko.
47:10
SPEAKER_01: Thanks to you for tuning in.
47:12
SPEAKER_01: If you like the show, please like and subscribe andor rate and review it on the platform of your choice or platforms of your choice.
47:21
SPEAKER_01: We also publish a sister newsletter every Thursday.
47:26
SPEAKER_01: You can find more information about that and the podcast over at automated.fm.
47:31
SPEAKER_01: Thanks again, and we will catch you next week for another episode of Automated.