
The Third Angle
Best Business Podcast (Gold), British Podcast Awards 2023
How do you build a fully electric motorcycle with no compromises on performance? How can we truly experience what the virtual world feels like? What does it take to design the first commercially available flying car? And how do you build a lightsaber? These are some of the questions this podcast answers as we share the moments where digital transforms physical, and meet the brilliant minds behind some of the most innovative products around the world - each powered by PTC technology.
The Third Angle
isento: A robot for every household
“We are on the brink of a robotic revolution. In a few years it will be possible to have a [robot] that is a reasonable member of a family or household”
The dream of robotics is to take on the mundane tasks of everyday life, so we can focus on what’s most important to us. And with developments in artificial intelligence happening at such a rapid pace, that future is drawing ever closer. So what’s needed to bring AI and robotics together in a way that’s truly transformational for humanity? When will we have our own Rosie The Robot from the Jetsons?
IT solutions company isento GmbH is opening robotics up to everybody with pib - their printable intelligent bot. Pib is a 3D printed robot with a humanoid face, moving arms and a torso - but the hope is that one day it’ll be so much more. We meet the visionary behind pib and CEO of isento GmbH, Jürgen Baier. In this episode he explores the features of pib, and how he’s enlisting the help of the global community to achieve isento’s bold ambitions for the project.
We also hear from Jon Hirschtick, who heads up PTC’s Onshape division. He explains the benefits of Onshape’s ease of use technology in isento’s mission to make pib accessible to all.
Find out more about pib here.
Find out more about Onshape here.
Your host is Paul Haimes from industrial software company PTC.
Episodes are released bi-weekly. Follow us on LinkedIn and Twitter for updates.
This is an 18Sixty production for PTC. Executive producer is Jacqui Cook. Sound design and editing by Ollie Guillou. Location recording by Liew Niyomkarn. Music by Rowan Bishop.
Welcome to Third Angle, where our robotic future is in everybody’s hands.
I’m Paul Haimes from industrial software company PTC. In this podcast we share the moments where digital transforms physical, and we meet the brilliant minds behind some of the most innovative products from around the world, each powered by PTC technology.
Artificial intelligence has come a long way – just look at the waves being made by AI chatbot ChatGPT right now. But for many of us, we dream of this world where AI is not bound to a computer screen, where a world of smart robots ready to do our bidding take care of the mundane tasks of everyday life. What would you make your robot do? You might think such technology would cost a fortune. But PIB is putting robots in the hands of everyone. An acronym for Printable Intelligent Bot, not only can PIB parts be printed at home using a 3D printer, but it’s also an open source project, meaning that anyone from anywhere in the world can study how PIB is made, make suggestions, and even modify it.
Isento, the company that created PIB, has been humbled by the growth of the project. It’s now even being used in schools to teach students. Jürgen Baier is CEO of Isento and the founder and visionary behind PIB.
So this is our robotics lab. When we enter, we see on one side the 3D printers that are working on producing many parts for PIB. And when it is turned around we can see PIB, the main actor now in this podcast.
PIB is 3D printed, which means PIB consists of plastic parts. We use PETG as material. And usually we use white PETG. That gives it a very friendly and a little bit of a sci fi look. We try to keep the size of PIB as close to human proportions as possible. Actually the shoulders of PIB roughly fit my shoulders. That was the easiest way to measure some stuff and to implement it that way. So far, we have the plastic parts that set up the complete upper body. And since we don’t have legs and stuff, we have a mount, which is just a bowl that is supporting the chest of PIB, and gives us an easy access to all the electronics and mechanical parts, and also makes it easy to transport and to bring it to a fair or something.
PIB can currently perform basic movements, like lifting an arm, or stretching an arm, open and close a hand, or also moving the head. Additionally, we implemented some more complex skills. One is that PIB can imitate the hand movement of a person that, for example, stands beside it. We use an external sensor, and if you open and close your hand, PIB follows this movement. And you can also move just single fingers and PIB tries to imitate your movement. Another a little bit more complex skill shows a social or curious side of PIB. We are using basic AI technologies so that PIB can detect, and also follow, with his head or vision, a person standing in front of it or passing by.
So what we are currently doing is we try to work towards a finished state, a stable version of PIB, and we tried to finish this in February. So this is also why we are currently here in the robotics lab, very busy, and assembling and disassembling PIB several times a day to test new parts. And currently you can see the arms laying down on the table. I will lift one arm, and what I’m holding here is one arm and one clavicle and it weighs about 3kg. Actually, it’s the right arm and the right clavicle. And what we are having are the motors that give the shoulder movement. So this is exactly the same movement that we have. We have upper arm movements, elbows, lower arms and hand.
The face actually was one of many things we thought about. We tried to give PIB a very friendly look. So this is why we chose a humanoid, or a human face. At the same time, we tried to give it a quite neutral look. So we thought, if it’s always smiling, that might look very friendly in the beginning but after one or two days, if PIB is always smiling at you, it might give you a bit of a strange feeling. So this is why we chose a neutral look. And right now, actually, we are updating the face. PIB is a community based project, so we try to get as many people in as possible, and we try to listen to their opinions. And one of the topics that people brought in was that the face may be even a little bit too human. So, for some people, because we have cameras in the eyes, the look was a bit scary or spooky. So we try to listen to that. And we are about to develop a face that is a little bit more abstract that creates a maybe a little bigger emotional distance between the user and PIB itself.
The biggest change will be, I guess, the nose. Right now it has a human nose, which sticks out a little bit. And we are going to use a camera that has three sensors. And this is quite a challenge. We have two sensors at the positions of the eyes, which is like a stereo camera, a standard tool that that is very often used. But also we have a depth sensor in the camera. And this is a third one, that would be something like a third eye. And we tried, and having a third hole between the eyes is giving it an even more spooky appearance. So this is also something that drove us towards making the face a little bit more abstract. And now, this is maybe a little spoiler, most probably the abstract face will have a stylized nose that will consist of three holes that are arranged in a triangle. And the upper hole will hold the third sensor of the camera, and the other two holes are just for aesthetic purposes. So this will depict the nose as a triangle. And that might be the biggest change. The mouth will be quite similar, less human, a little bit flatter. So the lips are not that prominent as they are now. But this will be the most prominent change right now I think.
Actually, this week, we achieved a major breakthrough. One of the parts that was bugging us for quite a while was the thumb. The thumb is a quite difficult part of the human body to mimic. And we had some difficulties to create or to construct a thumb that allows a nice gripping movement. So, in our assumption, gripping will be one of the first tasks that people will try to use PIB for. And so the thumb opposition is a feature that is very important that will be very prominent, and we tried to implement a smooth and soft bionic movement. Bionic means as close to the human body as possible. And I think we achieved something this week that helps it a lot. Also this is why the arm is detached right now. Because on the 3D printers, you can actually see the fingers coming out, we have four fingers on one 3D printer, and we have the scaffold that attach the control wires for the fingers to the upper body on the other printers. And we are trying to improve, build, test, improve by hand in a cycle so that maybe at the end of this week, we can say hand is stable and this can be published like it is.
Constructing a humanoid robot is a big task. Isento, at the moment, has 65 employees. It’s very clear to us that this undertaking, this project, is too big for a company of our size to do it from A to Z. It’s just a huge, immense task. For us, we strongly believe in the capabilities of a community to create large product to realise visions that people individuals have. So it was a very close thought for us to open source PIB, to build a community that people can join in, where specialists of different disciplines can join, exchange, and foster the development of a common idea, of a common vision, towards the humanoid project. And also, we want PIB to be an expression of how we see interactions, of how we see communities, of how we want to make our footprint in the world, that we have a product that is carried by a community, and developed by a community, that it connects people, so that it’s not only a robot that can be assembled and then that can be bought, but it’s something that people can join in as part of their life and where they can be creative. And the outcome of the community ideas and of the community contributions already have a great impact, and they change a lot. PIB is much better already than it would have been if it was just us. So the community is, for us, the way to tackle such an immense project.
When we started the community part, when we published PIB in the open source, we didn’t really know what to expect. And the acceptance and enthusiasm of the community really surprised us. So people joined in, in flux. So we have much bigger community than we dreamed of. And people really come up with great ideas. And actually, the current shoulder version, with the implementation of the current two degree of freedom shoulder, this is a suggestion that came from the community. As I said with the thumb, this is quite a critical part. And the shoulder is also a critical part for the arm to allow a free movement of the arm. And we tested many different concepts for that. So we discussed many different concepts internally, and the one that is currently implemented actually is a suggestion from the community. Then we also had, in the very first version of PIP, we had the motors for the finger very close to the fingers, which seemed for us a very smart way to do things. The problem is, when you have the motors in the lower arm, the arm gets very heavy. So we had one thing done very well, the fingers were moving nicely. On the other hand, the arm was very heavy, was very difficult to lift. So this is also an idea from the contribution. We have a wire concept right now. So all most of the motors are located in the upper body. And the movement is transferred through wires, which makes the arms much lighter and which gives us much more freedom to move some things around.
Looking back, we questioned ourselves why we didn’t start with the idea right away, because it’s so straightforward and so logical. But we were just following our thoughts. And the suggestions from the community had major architectural changes on PIB right now. So if you’re listening to this podcast, and you’re interested or excited by the idea of a humanoid robot, you can join our Discord server, you’re very welcome to join there and say hello, and try to follow what we are doing, get involved in the current processes, and we would love you to contribute and to become a valuable member of our community.
We are on the brink of a robotic revolution. Our daily lives, work lives, and also private lives will be very influenced in the future, or in the next few years, through robotics. Currently, we have simple helpers at home. We have dishwashers, washing machines, we have a device that helps us hoovering our living rooms. Some people have stuff in the garden that cuts grass and stuff like this. My firm belief is that in future there will be a strong development towards more generalised support robots that can take over daily business tasks. They can help us at home, in my personal dream they can find my keys, they can help me to make my morning coffee, and when I come home, the kitchen is already clean and everything is in its place. So these kinds of tasks. I think this is a big movement, and we want to be part of this movement. And actually, we do not only want to be part of this movement, but we want to be at the forefront. We want to lead and we want to shape the development towards this idea. The time until PIB is really able to take over many household tasks depends a little bit on the community. So we do our best to push people forward, but we are not able to do the whole task. So if people continue to join our community and contribute the way it is, I would say in a few years, it will be possible to have a PIB that is actually a reasonable member of a family or household.
That was Jürgen Baier, CEO of Isento. There’s a wide variety of skill sets brought together with the PIB community project, and not everyone has experience using computer-aided design software. And that’s why PTC is CAD solution, Onshape, is the ideal solution due to its interface and ease of use technology.
There’s really three things that make Onshape ideal for the PIB community project. One is everyone can access it easily. In order to use a system, you have to be able to use the system, you have to get to run on your computer. Onshape is the only professional-level CAD and PDM system that is truly cloud native – it runs in any web browser, even on any mobile phone or tablet. And so what it means is everyone on the team can get access to it instantly. So a new person joins the team, as long as they have a web browser, and everyone does, boom, they’re into the system.
Point two is all the product specifications, the models that drawings, even other documents, and information in the Onshape document, that can all be shared in real time with everybody. There’s no copying, there’s no “Who’s got the latest version?”, it’s shared in real time in a database. So that means that everyone can collaborate and see the up-to-date model, or they can go back to any previous state – there’s even branching and merging, like you see in collaborative software projects. So that’s point two, all the data shared in real time, no files, no copies.
Point three is the change capability. Anyone on the team can fearlessly make changes and explore design ideas without worrying about overwriting each other’s work, without worrying about messing up the system. Because again, change is built into the fabric. We have a next-generation approach to PDM, to data management. And that means the team has opportunity for all this creativity while still having the data under control and management.
And with the large number of contributors to this project, Onshape’s database-driven architecture comes into its own, since it allows you to track who’s done what, to the various design files. The more people you have on the project, the more Onshape is valuable, because we track every edit that’s ever been made, who did it and when. Not every version of a file or an undo list or something, but every single edit. So you can see what people did and track it all. It makes it really easy for this kind of a project where lots of people are collaborating and you don’t have to worry about losing data. You just can’t do that with older file-based tools.