114: ‘Technological Telekinesis’, with Dr. Madeline Gannon
A conversation with Dr. Madeline Gannon.
Dr. Madeline Gannon of ATONATON joins the podcast to talk about the potential of robots and machines as they relate to humans, the difference between programming robots to do a task versus creating an interaction model (AKA how to talk to machines), how working with robotics has led to what she describes as “moving matter with our minds”, her next endeavor with cable driven robots, her work on the continuum of creating intelligent and autonomous machines, and more.
Latest TRXL AEC/tech Newsletter
- ATONATON website
- ATONATON on Twitter
- Madeline on Instagram
- Madeline on Twitter
- Madeline on LinkedIn
- ATONATON’s Googly Eye project
- TRXL 028: ‘A (Robotic) Holiday Spectacular’, with Brian Ringley
- TRXL 059: ‘Fundamentally Changing the Equation’, with Brian Ringley
Watch this episode on YouTube
114: ‘Technological Telekinesis’, with Dr. Madeline Gannon
Evan: Welcome to the TRXL podcast. I'm Evan Troxel. This is the podcast where I have conversations with guests from the architectural community and beyond to talk about the co-evolution of architecture and technology.
In this episode, I welcome Dr. Madeline Gannon. Madeline is a multidisciplinary designer, blending techniques in art design, computer science and robotics to forge new futures for human robot relations. Also known as the robot whisper, she specializes in convincing robots to do things they were never intended to do. From transforming giant industrial robots into living, breathing mechanical creatures to taming, hordes of autonomous machines to behave like a pack of animals.
She believes that technology is a cultural medium and tunes her work to engage communities across science and society. Her works have been exhibited at international cultural institutions [00:01:00] published at academic conferences and profiled at global media outlet, such as the BBC, The Guardian, Financial Times, The Science Channel, Wired, Fast Company, Dezeen, and The Verge.
Madeline is a World Economic Forum cultural leader, a former robotics and AI researcher at Nvidia and a former artist in residence at ETH Zurich, Autodesk Pier 9, and the Carnegie Mellon STUDIO for creative inquiry. She is known as one of the top 10 women in robotics industry. And world's 50 most renowned women and robotics, according to analytics insight.
Madeline holds an M.Arch from Florida international university and a PhD in computational design from Carnegie Mellon university. She leads ATONATON, a research studio inventing better ways to communicate with machines where they combine the innovation of a research lab with the ingenuity of a design studio to build functional prototypes of alternative futures. [00:02:00] They are advanced Scouts for what normal may look like in 10 to 15 years.
In this episode, we discuss how metal and properly misuses technology in the most interesting way possible, the potential use of robots and machines, as it relate to humans. The difference between programming robots to do a task versus creating an interaction model. AKA, how to talk to machines. How working with robotics has led to what Madeline describes as moving matter with our minds. Her next endeavor with cable driven robots, madeline's work on the continuum of creating intelligent and autonomous machines, and of course Madeline's favorite movie robot must be revealed. Buckle up folks. This is a good one. So without further ado, I bring you my conversation with Dr. Madeline Gannon. Madeline, welcome to the podcast. Great to see you.
Madeline: Thank you for [00:03:00] the invitation.
Evan: Yeah. I'm happy to, have this conversation with you because of the amazing work that you're doing with robotics. You were just telling me that you're even kind of switching gears and, and learning new things, which I assume is just kind of par for the course. In robotics, you're basically blazing new frontiers with human to robotic interaction.
Maybe you can catch us up on. How you got to where you are now, an introduction to the audience of your work and the trajectory that you've been on,
Madeline: Yeah, certainly. I started off in architecture it's always exciting for me when I get a request from, a podcast about aec. like, oh, they still think I'm an architect. This is amazing. All right.
I've been doing with robots has been, it's so fringe, but my heart, my feet are planted firmly in architecture.
Madeline: It shapes the way that I see things. It shapes the personalities that I give these machines. but basically I've, built a career out of, properly misusing technology in the most interesting way possibly. So, started off in architecture, and even [00:04:00] my, my PhD work is actually from a school of architecture, although you would, never guess, you would think it's from a robotics department, but, my wander from traditional architecture towards robotics really came through the path of computer aid design and computer aid manufacturing.
so I, I came out of school during the recession, the 2008 recession, and there were no jobs, nothing. And so I, I decided to continue to study and to specialize in something. And that same time, this whole world of computational design Was so cool. Grasshopper was just becoming a thing.
All the work of some of the architects who I admired were as doing pric and gorgeous stuff. All stuff I did not know how to do. And so, decided to continue study. I I moved to Pittsburgh, went to Carnegie Mellon learned how to talk to machines. one of the amazing things about Pittsburgh is that it just so happens to be the robotics capital of North America.
there's more roboticists per capita there than anywhere else in the us.
You go to a coffee shop and you hear people talking about like[00:05:00] oh I am trying to figure out how to dispense this for this gasket on the store and can't quite get it Like that's normal coffee shop talk which is pretty wild and so even the artists that are there are actually roboticists that's where I sort of found my niche I was greatly inspired by all of the amazing work in architectural robotics being done in Europe at the ICD Street Guard et t h Zurich but I was just one person one individual and the department there that did have a big giant robot and no one knew how to use it It was just sitting lonely in a cold dark third sub-basement below ground And so what was nice about that is that you know I knew as much as the teachers was basically nothing out of a lot of PDFs over and over and over again and sort of figuring things out through trial and error I start to be able to use these machines and with the secret superpower of Carnegie Me programming I sort of saw that I could fill a niche where you know I'm not gonna build [00:06:00] pavilions with these machines but I can explore interaction I can explore interfaces I can explore how we use them in a more intuitive and engaging way And that was a territory that a life of its own after a while And I'm still kind of riding that wild pony uh into the future
Evan: So you are the robot whisperer it's interesting to hear the way that you talk about the robot You know you said the the was lonely down there in the sub-basement and it's interesting to me to hear somebody talk about a machine like that And I think a lot talk about machines like that but not necessarily in architecture right You've got woodworker and people who use C N C and in like they're intimate with their machines that enable them to do the crafts that they do very similar to that I assume So how do you see that Because it is interesting to hear you talk about them that way
Madeline: there's a spectrum Yeah So you know you can go from appliance your washing machine That's just a robot Yeah then you have something like your car that's somewhere in [00:07:00] the middle where you it has its own personality meaning you give it a name you connection to it it's a part of your identity And then on the other side of the spectrum maybe you have your pets also these non-human things that you cohabitate with that you can't really talk to but you get a lot of value from their pure existence next to you And so I think robots sort of can occupy part of that spectrum between car and pet In
exciting way than you know just doing things that are and efficient and automated
Evan: It's interesting to hear you even talk about a washing machine as a robot cuz I've never thought of it that way I think of it as a machine But robots are machines It's like a categorical maybe slight nuance to to all that I've owned a lot of cars I don't think I've ever really given my cars names I think I've owned like 38 cars It's a lot of cars Okay but only two robots And those are both what I consider robots I've had two vacuum robots I just recently got a new one that does the mapping And draws straighter lines than the old one used [00:08:00] to important as an architect that that the lines on the carpet are are nice and aligned with
it is interesting to think about machines doing these tasks and Brian Ringley has been on the podcast before and he's doing work with SPOT at obviously you know Brian well and I recently heard on a podcast about the utility of these machines to do things While we're sleeping Especially with the work that he's doing in construction tech with spot going out and doing sidewalks and scanning bringing that data back But the stuff that you're talking about is again like on a different part of that spectrum and you're talking about human toro robot interaction and you know one of the projects that I've seen you posting on social media is the big cka Is it a Cka I think the big robot with the big googly eye on it And it is amazing how what seems kind of like a silly little thing makes it way more I don't know what's the word Attractive the way that you interact [00:09:00] with it it makes it more human or less maybe um I don't know it it it just seems more approachable is is that really the root of your work there Is is because the utility of robots is unmistakable but there is this hurdle That we have to get over as humans everybody wants to kick the spot robot away from it and see what it does Right Even one of the latest episodes of the Mandalorian like where there's the drones walking by the Mandalorian is is kicking drones just like people would push on the Boston Dynamics robot and hope it doesn't turn around and Right Kick back and you're breaking down those barriers with your work in robotics that that's really what it seems like
Madeline: me I I see that we're in this era Of technological telekinesis Right our technology is advancing so fast that we're reaching the point of
supernatural powers And maybe it's a framework of magical realism Maybe it's a framework of telekinesis but the idea [00:10:00] is
that we can move matter with our
minds just with
interfaces with the Animate machines that are roaming around
it's interesting how quickly the technology normalizes So you know you think of a washing machine Oh I've never really thought about that as a robot you call your vacuum a robot vacuum right
Madeline: have a robot vacuum too but she's gonna grow up just calling it a vacuum Like
Madeline: the distinction won't be there for her at all
and we think about that too Like there are more robots than we imagine
Madeline: there to be in general Like
the Super Bowl you know there's robot cameras that are flying around getting
those shots Rihanna's halftime show you know those platforms those were cable robots moving things
up and down and all around as soon as it becomes normalized we don't acknowledge it as a piece of technology anymore which is such a beautiful part of how our our brain works the project with the big giant googly eye [00:11:00] to take an idea its maximal and a little bit absurd but also recognize just the power of our brains to render life onto bare bones Geometry
Madeline: just a big two meter white circle a one meter black circle And as soon as it starts to move we see this creature It's no longer an industrial industrial infrastructure now this like otherworldly thing floating in front of you
Madeline: uh perhaps a little too pushy and uh attracted to you
Evan: Well maybe you can make the distinction between that and animatronics I think a lot of people have interacted with animatronics which maybe you would consider those robots as well I mean obviously there's crazy stuff going on on like the Disney side of things At the Imagineering you've seen the Spider-Man Swinging robot that lets go of the web and does a flip and it's incredible to watch and even how that's kind of navigating environmental aspects in real time so that it's compensating [00:12:00] for It's flight path and the way that people have interacted with animatronics with expressions in facial robotics it's getting absolutely incredible the amount of nuanced communication that happens through that then you've like the big googly eye and how much we can draw from you said it's it's kind of pushy sometimes Right But you're you're basically feeling a feeling from what it's presenting to you and and you're also driving that through interactions with your body language Is that correct
Madeline: Exactly Yeah what's amazing about our lower level brains our mammalian brains is that it receives these kind of low level frequencies that we just can't turn off just our our instinct to begin to analyze things as soon as they move is this is this friend or is this phone you know is this going to eat me Should I run away And that is I see such an potential interaction medium
for working with these machines We're working with technology in general for AI systems [00:13:00] robotic systems because you can begin to pierce attention just by doing these low frequency gestures for animatronics a lot of times they wanna do high resolution They you wanna see eyebrow move in 16 different ways and they're going for realism and that's amazing Disney research is one of the leaders in the world in robotics research They do amazing amazing things of times because they have the most interesting problems right It's driven their movies It's driven by their theme parks
Madeline: really great place to work too but that's not the only answer I guess is what I'm trying to show
Madeline: a a lot more accessible ways nuanced ways still connect through technology
a more contextual way by doing things that are just tuned and sensitive to their environment maybe that's where training as an architect comes through I see it like go to architecture school and the gift that you're given is a hypersensitivity [00:14:00] to how people move through space
Madeline: learning how to talk to robots Like all of a sudden I could give them those same nuances and those same sensibilities and have them sense and respond to people a way that I guess is just a lot more lifelike and a bit more has a a breadth of consciousness we don't yet always see our systems our technology at all
I think we're heading in that direction think I'm just a little early there
for me at least that's a way that I would like to with technology where it's beginning to anticipate what I like than me having to command it
Right Yeah It's interesting to think about kids right I have a 17 year old and the Way that you see them mature over time Going from waiting around to be told what to do to figuring out what needs to be done and being a useful helper Right And I think that's kind of what you're talking about right this noticing just watching this is another thing that I think we're getting more and more accustomed to but maybe [00:15:00] aren't completely aware of like our phones They're always on Even though the screen is off the machine is still on And we see this in our kitchens with an Alexa device Right hopefully nobody's Alexa device just went off when I said that but they're always listening So they're always listening They're they're attentive in that way because there's kind of this wake word or you know it could be a motion it could be anything This is becoming more and more prevalent even if we're not aware of it but we still expect it to work all the time And so there's that and then there's this idea of sensors in buildings iot that are always watching They're watching you know for movement They're watching for sun hitting the glass Does it come like should I lower the shades in this kind of proactive stance of just watching and listening and measuring things as they're happening in real time so that they can proactively do something on our behalf all of these things tie together right This is all and I guess what I want to know because you said your feet [00:16:00] are still firmly planted in architecture and the work that you're doing is with this interaction How does that tie together do you have a big vision of how you see that tying together Or is this just experimenting and figuring it out as you go
Madeline: Oh I go by feel is no vision tend to spend a lot of time research I have these cycles where I'll do a research cycle a production cycle then a promotion cycle And that seems to be the rhythm of how I work and investing a lot in research and being able to set a good North star that becomes the heading then I just wander towards that So I have confidence in the direction that I'm going I don't quite know where I end up
Madeline: that's kind of the joy of working in research is that your job is to be an explorer
Madeline: is to find the unknown report back What's interesting about it
Evan: I'm writing down some notes here as about some of the you're talking about here So when you set a direction and you're heading in that direction with your latest work said you don't necessarily set where you're going so what [00:17:00] are some of the surprising things that have come up this latest round in this latest project
Madeline: So it's early days I'm doing a new project that is trying to create open source for cable driven robots and cable driven robots is a relatively new area of research and robotics about 10 years old a really nice sweet spot because a lot of work already done by that are much smarter than me in terms of control software the fundamental technology but there's a lot of overhead for creativity and interaction that because the community is so small arts and design have not yet come in to really showcase the full potential of what it can do And so that to me is when when I start to get goosebumps it's like okay there's something here There's something here don't know quite what it is yet but it's gonna be at least valuable to a couple different communities and so what I'm often surprised by is how much I can find on the internet then on the flip side of that is also how much knowledge my friends have balancing that with you schools [00:18:00] on a brand new subject matter up to speed very quickly and then reaching out to my community of friends to To really get deep wells of expertise to sort of get things kicked off quickly is always a it's been a lovely way to operate in this world
Evan: What do you think the advantages or what what do you see the potential being in cable driven Than robots
Madeline: if you think about cable driven robots are basically a big Winch It's a motorized winch a spool of wire a cable that you can it to unspool and then spool back And when you connect two of them together now you can move in 2d And four of them together Now you can live in 3D and there's crazy if you really wanna get nerdy into this uh you can go on YouTube and look up cable driven parallel robot C B P R and you'll see some that have people inside as the end effect inside the end effect getting whipped around the room very high speed very precisely And so what's amazing about them is that you know how do you do an installation with robots that fills a courtyard or [00:19:00] an atrium an arena
you know that googly eye project that I did that has the biggest robot that ABB makes it
Madeline: 10,000 pounds It's 14 feet tall to its shoulder so it can reach like don't know five or six meters tall It's it's standing in front of a Toronto source Rex I can't scale up any bigger than that with
Madeline: but all of a sudden I just get a bigger spool with bigger cable attach that to the same motor now I can span a wider area and bigger things and move them faster I think there's a lot of potential there to begin to explore scale immersive experiences that don't have screens are really about physical world and connecting with our physical world And
Madeline: that's also my bias as an architect right I wanna bring of the magic of the digital Into our tangible existence
Evan: Yeah I'm getting off of the screen into 3D space right I was thinking as you were talking and then you nailed it there right It was like this idea of performance and Maybe not just even [00:20:00] entertainment but I'm sure that's obviously an aspect of it And that really gets back to even the work you're doing with the googly eye right It's the way that this robot moves spatially and to get back to your earlier comment about being about how people move through space and this starts to become This really interesting interaction between space and humans and machines all at the same place at the same time I can see that just opening up new worlds for this kind of thing right a lot of people don't expect that but I think again we are seeing it we see it with Super Bowl halftime shows We see it with the fireworks drone displays where
even there that might feel like a projection on a screen but it is happening in 3d It's like the first time that I heard that constellations are not 2d they're 3d and it's like okay that seems obvious after you say it but until you say it it's not obvious It's just a projection on a screen and so I could really see some interesting outcomes From that type of interaction and that kind of work And it seems [00:21:00] like you're driving you're designing interfaces to drive these performances human movement potentially or you know again the robots are watching for us remember watching a video of you with the googly iRobot right using your hand to basically
guide it and tell it what to do based on a gesture in space And so there is a lot of really interesting potential there can you speak more to that side of things and like that interface design I assume that's where a lot of your research actually is is can you start to communicate with the robot that it's not through code and it's not even through verbal communication but it's through these gestures
Madeline: Exactly like for me Interface is everything the technology
Madeline: I sort of relearned that lesson with chat G P T coming out
Madeline: So you have a large language model that was about a year old the technologists developing it were so underwhelmed with its performance because they had latest and greatest yet to come But all of a sudden you have a good interface that lots of people can [00:22:00] use and it takes the world by storm that's the same lesson that the iPhone taught too existing technology but put into stable hardware good software interface is everything And we haven't quite reached that yet with robotics And we also think of robots as these things like individual entities like the Boston Dynamics robot or
Madeline: adjuster robot arm But we're at a potential where can begin to think about whole building systems robots We can think about large areas that are aware of what we're doing You know when you can begin to sense and respond and manipulate the world that's a robotic system And it may not need to have a physical presence that we normally think of as as an individual entity right It could be something that we're already amongst us and already inside us that has a bit more ephemeral physical form I see my role as in the research that I do as reaching towards the pragmatic and the poetic And so mediums that I work with very very pragmatic They're engineering platforms both in the software and the [00:23:00] hardware And my role is to discover all the good ideas that were left on the cutting room floor That maybe were discarded years ago and haven't been brought up again or that know the low hanging fruit that because of your siloed background you might not even see And I enjoy that a lot with just also helping show bit more of the possibilities and preferable futures we'd like to have with these technologies for our society We often Have a good idea collectively of the futures that we don't wanna have with AI and the futures that we don't wanna have with robots There's endless science fiction and movies
that but the futures we do want to have with them are a lot more nebulous and they've really yet to be defiant so moving to the future I want my work I aspire for my work to begin to offer this Optimism and hope and amazement at the time that we're living in which I am so grateful to be here And Janelle and by sort [00:24:00] of projecting into the future I feel like I can bring it a little more close to the present my work
Hmm Yeah Interesting the whole idea of this potential is so interesting to me and thinking about it as an architect you can list a few uses that you see architects as we are planning buildings that will be operating when achievements that you're working on will be at some further level down the road really not knowing what those are but kind of forecasting what do you see the potential there for architecture and robotics together that people might just want to be aware of now that just start thinking about those and start igniting the get people's the architecture side gear turning about that And that could be architectural technologists That could be actual building architects It could be teams that combine both of those people What do you see happening out there in the future
Madeline: What's happening now which is very exciting for me I think is what's happening in construction robotics So that is where a lot of the direct the Venn diagram from
advanced robotic systems and [00:25:00] AEC are happening right now
for example using autonomous diggers to do site work
Madeline: and that's again something that can run 24 7
Madeline: and can begin So you do your site prep and your site is prepped by the time you get to the site
Madeline: fantastic there's a lot of work in architectural robotics that for the past 10 years 15 years it's really just been about discovering what we can do with this technology And the next 10 to 15 years is really beginning to bring that out into industry and apply it in a meat and potatoes way in a hard hard engineering way it's exciting to see it I don't wanna say mature but I I'm excited to see it expand outside of architects talking to architects about architecture
for the futures that I hope to have we have environments with robots roaming around Like hospitals are a great example tug robots that are delivering non-essential things between you know central storage and an individual [00:26:00] patient's room or warehouses where warehouses are employing more and more people every day And so for those interactions to be for everyone involved would also be great there's already places where it's very normal to work alongside a robot and just having a better experience I think can just have an immediate impact today I don't know how architects begin to play into that but I think that there's the design of space and For a long time I think the architecture discipline has been focused on the formal design of space the construction and and built environment with that space but not necessarily the experience of that space phenomenological So the atmosphere that it's created I think in general it's largely ignored this ubiquitous medium of technology that exists in every volumetric inch of every space that we're in Now
and there's just a huge potential with learning a bit more of the technology side of things [00:27:00] for architects to be able to claim that medium creative medium and add that as a layer into how we experience spaces
Evan: can you give some examples of what those are we hit a little bit of the iot internet of Things kind of sensors and things like that that are reading that But what else is there
Madeline: Hmm I'm trying to think about how crazy versus how pragmatic I should be here
Evan: can we do
Madeline: We'll start with pragmatic use cases we have a built environment have these office buildings Office buildings are at 35% occupancy annually And so how can we begin to occupy that space more efficiently Okay well maybe we have self reconfiguring spaces that
Madeline: to change and adapt as we have them for these different things I think a world of scarcity that we're marching towards doing more with less is a thing that reconfigure environments and reconfigurable reusable spaces is automatically useful for And then maybe the more delightful things is when you walk into a [00:28:00] supermarket the door automatically opens for you
don't think about that as a
Madeline: robot that welcomes you to a space but you know but it is
Evan: has potential to be welcoming
Madeline: blast of cold air
but you know I'm working now in Miami and we get the sporadic brains and how nice would it be if the building that you're in comes and covers you as you're trying to cross from one building to the As you cross a courtyard for example Right And that's one of the things I'm thinking about with these cable robots and and spanning a large courtyard is can we make shade for someone who looks hot
Madeline: the little things that that begin to take subtlety nuance that we're taught to appreciate and work and use technology to add that layer into the experience of space
Evan: I think where most architects are met with that is that the idea of the first cost versus The [00:29:00] cost over time of a thing and clients are typically focused on not all clients but a lot are focused especially in the work that I come from which is public work it's like what's the first cost That's the only thing they care about right now somebody else is gonna move into administration and they're gonna figure out the budget after they're gonna even have to figure out how to maintain the building cuz that's not even part of the first cost And then the things that you're talking about Are very much not well they are a high first cost but the payback over the potential is quite a different story because talking about reconfigurable spatial robots right Potentially you're talking about things that provide value to the humans occupy that building over time and can be adapted over time to do More or better I mean there's a lot of different categories you could check a box there for it's interesting to me to think about this and are there examples of people who are doing things like this now that you're aware of
Madeline: well reconfigurable spaces for example there's a high-tech solution and a [00:30:00] low-tech solution but lot of the sensing to know when and how to reconfigure a space is the same whether you use a robot to move a space or you give instructions to someone
Madeline: just a modular kit that is expandable and adjustable
what's interesting about robotics that a lot of people don't quite know is just how much of it is really software The way that the thing moves is one part of it you can have software that impacts the world that can modify and manipulate the world by sending a text message to someone say Hey can you move that from here to there
that's an AI system or a robotic system That's the brain of it without the body
ways to there's you know blue sky dreaming about what we would like to see in the world
always a way today to make it happen compromise on certain elements of it Like if it's really a good idea there's ways to make it happen What I see now is in the industry you start to see experiential design groups studios within larger firms So [00:31:00] Gensler has an experiential design group Rockwell has an experiential design group And a lot of this is is like what you said you know they're working for retail they're working for casinos they're working for restaurants things where Instagram photo
Madeline: Gold for those clients That's where the value of proposition is going to differentiate you from the competition by making these incredible experiences for people that they're gonna talk about and share but this is just the beginning so new that there's not even enough examples for there to begin to see what the potential accurate domain space could be I mean I wanna build drawing robots that can draw murals on the size of buildings but a pragmatic application if that is actually window cleaning on
Evan: Hmm mm-hmm
Madeline: now of a sudden instead of sending a person up who can risk a fall have a on the edge of the the edge of the high rise that's a mop Roomba suction cupped to the side of the high rise And so for me I find it a boring application but that's often where the bread and [00:32:00] butter of a technology is And finding a way to sort of do both both have the pragmatic and seek the poetic I think is also we're trained as architects to do you know the building has to stand it can't leak no needs to fit the program But how do we make it special How do we make it spectacular do we create atmosphere and make it remarkable for the people who exist in it every day And I see that as a potential for technology again to to add as a layer into how we experience environments
Evan: I think we really could take a chapter from the Entertainment Playbook I mean this is where we see the innovation happening We talked about Disney earlier We talked about the Spider-Man robot but I'm even thinking of Well we even talked about the Super Bowl Where it's performance You're watching it happen from from afar experience that I had was in Las Vegas going to see a Cirque de Sal show where the stage was a fully articulating robot I mean it was it was part of the performance stage would lift it would [00:33:00] become a giant wall it was an amazing performance to watch because think part of it was just how unexpected it was Not knowing anything about the show before we went to see it And then when you see it unfolding in front of you that story that's unfolding in front of you that's fully enabled by this fully articulating stagecraft performance piece robot Right I think it's really interesting now to think back that that was a robot doing that Right It was like the robotic arm that you're working with but 10 times bigger entire stage
carrying people I think the performance was called ka I don't know if it's still playing there but was like arrows being shot from behind the audience from performers and not really but you know that's what it felt like And they were hitting this stage And so these arrows were coming out of the stage instead of going into the stage from underneath And then acrobats were swinging and playing off of those arrows that were sticking out of the stage because the stage should be gone from being a [00:34:00] platform to a wall So not only is it supporting the people who were the performers they were swinging and flailing Like it was like the prices right Plinko machine of people Right And they coming down these arrows and falling and twirling around them and it was an amazing site to behold So get back to this idea of architects taking a chapter from the Entertainment Playbook affects so many things video games movies entertainment visual effects back to you know we can steal ideas from them because they have this recurring revenue of people buying tickets to see the things or to play the things or to do the things
Madeline: just charge admission to go inside our building
think a place where you see this as well is in airports airports uh public arts
Madeline: often have kinetic sculpture that begins you know it has a purpose It's to sometimes it's part of the way finding you'll have something that undulates to guide you sometimes it's there to calm travel can be quite anxious sometimes as
Madeline: and so that's where you start to see [00:35:00] these robot art begin to help create atmosphere
Madeline: and that's probably a more tangible everyday way where you actually will come across immersive robotics in your in a day-to-day interaction
Evan: well are we missing here maybe you could tell us a little bit about some of the previous projects that you even have done more recent history but before the Google ai like what has been kind of looking backwards and connecting the dots to robots now have you gotten to there
Madeline: So I feel like I've done Every possible thing on every possible robotic arm platform point And so for me I I'm ready for change
I stumbled into robotics by accident basically lot of the work that I had doing before my robot training series was about interaction design computer aided design So making the computer aided design and manufacturing tools that I wish I had
so I was designing software that let you 3D scan your body and drape really intricate [00:36:00] forms then using a adjust drawer interface to make for your body that you can 3D print and it fits you right away cuz it was built for your body and then taking that out of the computer actually projecting it onto your body So now all of a sudden you have a bracelet design projected onto your skin and a computer sensing system that sees how you push pull prod and poke that system to design it just the way you have And when you close your hand it gets exported and you can 3D print so the work that I did with industrials for the first time that was this was a part of my my dissertation at Carnegie Mellon The goal was to the system that let me you design on the body but I had to go send it to a 3D printer to then get it back and I wanted to attach a 3D printing medium onto the robot so I could design it And then as I'm moving around the robot is fabricating it onto me was doing that as a part of an artist residency at Autodesk When was this quite a long time ago at this point And they happened to record a video of me just debugging my system [00:37:00] and that hit the internet and took a life of its own And and everyone was just they had no interest in what I wanted to make and fabricate They were more interested in seeing this person standing in front of this beast of a machine that looked like a crusher but decided not to And me that was a revelation It was again know seeing the poetry in this when I was really at it from a pragmatic standpoint that the motion that this thing has that doesn't look like us can't talk to us you know has no way of really communicating just from the way that it moves people began to lead into it And what discovery That that's an entire interaction medium that no one had really explored before
Um not in construction robotics at all
and I think you know you go onto Hollywood animators and they're like yeah duh Motion expresses know feelings and
Madeline: You look at Wally Wally has barely has any
Madeline: it's just the motion and beeps and boobs that begin to pull at our heartstrings Animators have for [00:38:00] since the 1940s and we're just now discovering how to begin to apply that on our physical world
Evan: I love that you brought that story up because I even see it in your work now right When you post a video You're standing in front of the machine and it's doing something that is telling the story about this interaction the work that you're doing and it's so interesting But I'm wondering behind the scenes is it really like for you I mean I can only imagine how much time you're spending code to figure out How to do that So can you fill in the gaps of what we see on your posts versus what your research like how much effort what's that really like how are you blazing that trail there and I'm sure you see better potential in that as well but you're doing kind of the hard nitty gritty work right now
Madeline: it's been a long journey for that and so every iteration gets Better Every reation gets faster but every time I use a new machine it's like a different thing altogether If
Madeline: robot its presence different it moves different [00:39:00] It might be faster and more nimble and more versus the big ones that feel more like a lion roaring at you Cuz their motors are so big and
Madeline: around a bit so I try to be inspired by the individual machines that I was using the project in Zurich basically I've gotten by having really really cool friends have really really cool equipment and I operate in a very nimble way a very quick way I work a lot in simulation and it lets me move very very quickly so I can drop in to Zurich while they have a two week break in projects on their four ceiling hung industrial robots and do a performance with it and not be too disruptive to what's happening there I was really inspired by walking along Lake Zurich and seeing the swan sort of the tourist and beg them for food and so upside down robots had the kind of grace and demeanor of those swan necks and so the personality I'm rendering [00:40:00] into them was trying to replicate excitement and indifference that the swans can have for you as soon as you run out of bread but a lot of it is compartmentalizing so I put on my engineering cap develop these systems that begin to have continuous feedback between what you're doing and how the robot is moving And then I try to be more like Jane Goodall sort of stepping back and
Madeline: Observing how I working with them observing how others work with them and Then I can go back and tune the brains of the robots to begin to accentuate certain features or certain personality traits like curiosity or boredom or impertinence
Madeline: I find quite fun Those are all things you can do just with gesture which is is pretty wild and that's that's the general workflow of my motto is get it working keep running as quick as possible and then step back and observe And then learn from that and move again
Evan: I don't wanna belabor this but this idea of Programming robots to do something versus this interaction model are two [00:41:00] completely things and I think a lot of and students have learned what it takes to get a laser cutter to do what it needs to do or to get a CNC maybe to do and figuring out the tool paths ahead of time and programming the machine to do the task that may be repetitious It may be one-off whatever right Depends on what the output is versus this idea of interaction And so can you just speak To the difference there I mean because you're basically
there isn't because it's like behavior
presents itself differently
Madeline: like and maybe again like I approach this different than a lot of my computer science colleagues they think in math and I think in geometry And I think in tool paths what I'm doing is basically generating the tool paths as I go
tuning that trajectory to have more expression or less expression based on lots of iterations of doing this You know first project that I did in London Nemus I got to borrow industrial robot from an assembly line in [00:42:00] Birmingham in the uk to give it a holiday come and live in the museum And one job was just to exist and to hang out with people and to be curious and to be bored whenever people were boring around them And that has its own you know that was an early project and you can see it has its own like jitter moving around Cause I was still learning how to generate tool paths And it gave it kind of this quirky squeaky personality actually of the chirping of the motors moving in a jittery way then later you see that they're moving more smooth and more serpentine a bit a lot of that is because the way that I'm generating the geometry that they're following begins to have those same personality traits I'm just beginning to encode proprioception the way that our body feels in the space into Geometry that I'm feeding to the machines little by little little by little just like you would've CNC machine or 3D printer or a laser cutter those are three axis machines that go up down left right forward back robot is just a six axis machine It's the same exact thing It looks [00:43:00] a lot more complicated than it is but if you understand how to get a C N C router to follow a tool path you know how to program an industrial robot What the challenge is is that the knowledge of how to use them is oftentimes locked behind multiple PDFs and you just have to go and read the manual a lot of times
Madeline: again and then fail of times and then you reread the manual for the 16th time and it finally clicks what they should have written how to do it to make it un understandable and that's definitely getting better now You know I think maybe my first experiments with industrial robots has been about 10 years ago and now there's amazing platforms robots that you can program with Python or Blockly which is a visual programming language to begin to do simple tasks you're starting to see like robotics companies are waking up that need to make their stuff easier to use if they want to get more customers all lost our patients with the bad interfaces
Evan: Right I'm wondering a couple more questions [00:44:00] but we can wrap up But one of them is this idea of machines being available You talked about this machine it went on holiday it hung out with people in the gallery How much potential is there there's probably no way to quantify this but there are so many machines just lying around just waiting to be used going back to the early part of the conversation that lonely robot right It's just sitting there collecting dust in the basement
mean do you really see huge amount of potential It's the answer's obvious here I think But maybe something we could tease out of this with all of the machines that we have at our disposal the promise of a 3D printer in an architect's office is that it can be building models while you're doing something else right While something maybe more valuable maybe not don't really don't need to place a value on building models or not but it's the magic of it is not in the speed of it It's not fast but it's doing it While you're not doing it right The idea that Brian talked about was spot It's going through the buildings at night so that in [00:45:00] the morning you have information to make decisions based on what it captured and so the kind of untapped potential of all these machines that are sitting around us to do incredible things that we wouldn't have to do just seems obvious And yet at the same time it's everywhere Right
Madeline: I think And maybe this is again my bias always towards interaction but I sort of see the role of rapid prototyping machines 3D printers and laser cutters even robots As one aspect is that they can work longer and do things that you can work in parallel so
Madeline: to optimize for that but they also begin to remove all the friction from a crazy idea you have to getting something out in the physical world And I think that is really where a lot of the potential where interfaces like this come from So where the technology is really undervalued right don't have to think and wonder about something you can just go try it and know in a couple hours was a good idea or not or learn You can learn by doing a lot more than before We had this [00:46:00] technology for robots in general like the future that we are collectively racing towards is that machines will be autonomous intelligent and autonomous
by definition you know we're going from a paradigm of controlling them directly Programming Please do this then do that Now Do this you Go away to implicitly programming in They have the autonomy to decide where and how to move We're giving up that agency to them And if I'm sitting on an app and I'm ordering lunch and a delivery robot is going from the kitchen of a restaurant to my office that's a point Age B that's a transactional relationship to that robot is doing what a lot of robotics companies have ignored completely is that that Machine is operating in a public space in a public environment and it is interacting with tens or perhaps hundreds of people on its route to achieve a task You know robotics has a task-based model that is the main value set for how well a robot robot is doing as they come out of the [00:47:00] lab to live in the wild with us they also need to be good citizens good neighbors They need to be attentive You know there's so much of their existence that is outside of that one task that they're doing and we need to have good interfaces that make that a pleasant and beneficial experience for everyone
Evan: It does seem like there's a lot of room for delighting and sensing Opportunities to do things like that in that goal-oriented paradigm that robots operate under
Madeline: It could be you know the delivery robot goes and delivers the food and on the way back it picks up garbage that it happens to pass know there's this models of then maybe I'm not so annoyed at the fact that it just cut me off on the sidewalk cuz Oh it it picked up those cigarette
Evan: It all comes back to the vacuum robot and what it can pick up along the
Madeline: the perfect form of
Right All right so my last question and I think I asked Brian the same question I mean everybody who comes on who who talks about robots I have to ask who is your favorite movie [00:48:00] robot of all time
Madeline: That is very tough
Madeline: That's like picking your favorite child thing I only have one Um Wally
that's a good one.
Madeline: nonverbal communication.
Madeline: you know, the, the perfect anti-hero.
Evan: there's some incredible stat about that movie, and it's like the first I, 33 minutes, 30, it's some huge number of, there's zero dialogue in that film. And to your point about what is able to be communicated through bleeps and bloops and micro movements, like there aren't that many moving parts on the robot compared to like a fully, you know, a, a person.
Right? It's very different it's one of those. Storytelling, triumphs. it is absolutely incredible. I am right there with you. That's a great choice. Well, that's it for today. I think I, have really enjoyed this conversation. I've learned a lot and I am, uh, that we can. Continue this after your next project. I'm really excited to hear what you learn with cable robots and how that can [00:49:00] apply to the picture of AEC and the building industry and, other ways too.
I mean, I know that you're not solely, have your feet planted in aec, but that's not the only place your feet are planted. I, think so. I'm very much looking forward to where this goes, so thank you so much for having this conversation with me today.
Madeline: absolute pleasure.