Pushing the Boundaries of Robotic Development with Tiago

Watch this video to learn more about Tiago, a research robot being developed at Chalmers University of Technology to push the boundaries of development. Tiago aims to navigate its environment and adapt its behavior to it.



Transcript

00:00:00 hello and welcome to Charmers my name is Diago and I'm a research robot so we can like test functionality before we actually deploy it on the on the real robot person detected within one meter therefore I'm stopping the Thiago by power robotics is pretty much a test bench that you're able to use to test all kinds of different things and we specifically we

00:00:26 are testing na navigation along with the camera so for our case we wanted to be able to recognize certain objects and do certain things depending on which object it is I studied back in India uh so I completed my bachelor in electronics and Telecommunications engineering so uh the contents of the program I really loved about it so I applied and to my luck I got it I started with my automation

00:00:52 mechatronics bachelor's and after that used true some Masters and I thought that systems controls mechatronics was the most fun I thought basically like robots definitely are the future for this uh for this world so it would be really good to have a hands-on experience on a robot the biggest thing that happened like 2 weeks ago ago U randomly a single

00:01:15 file on the robot got corrupted yeah and because of that the robot was not able to navigate or start up properly and then you know it took us like 3 Days To Figure one day the diagnosis went one day exactly what what was the problem we figured out the next day we found out the solution implemented it finally to our luck got got it to work we can start with the camera yeah

00:01:41 it's an rgbd camera which means it's a depth camera and that is very useful because if we have depth then we're able to detect the the distance to objects in front of it the three Sona sensors one over here here and here so it could detect like if I keep my foot over here it could detect the distance between the foot and the Sona so that is one thing and the laser over here that is pretty

00:02:07 much the most important sensor because it's able to locate itself within the room that is in yeah and without that you couldn't navigate at all so with the sensor we're able to create the map of all the room it's in and along with the computer we can also say it to go to places within this map so in our case right now when it is a human it stops if it is also within 2 m of the human and

00:02:33 it can stop completely when within 1 M person detected within one meter therefore I'm stopping well at least what we work on is the navigation and able to detect which low room it is in so perhaps warehouses or it's able to navigate for specific things and it's also able to avoid humans which is very important maybe uh improvise a bit further with the uh gripper

00:03:00 like grasping onto the objects more firmly which we have not yet implemented so far could also work on the room recognition so it's able to like see if it is a table a chair then perhaps it's in a living room so you have the the basic stuff of actually being able to like interact with your environment chucking objects around actually grabbing

00:03:22 them so uh my name is Jasper I create and maintain the digital twin so to speak of Thiago so we can like test functionality before we actually deploy it on the on the real robot so this is mimicking how the the robot would actually interact with this environment picking up up stuff and maybe even walk around and actually do interact with the environment the good part about doing

00:03:50 this in the digital twin you can just there's no no risk of destroying anything uh in general we need that safety when we actually deploy stuff on the real robot we we need to know that it's not going to make anything unexpected we don't want it to run into the wall it might damage other people it might damage itself um and this scene we have it's the it's the old lab

00:04:15 environment right now we're in the new one uh but we will also have like a kitchen scenario uh which would be a more like real life scenario where we actually want to train a robot um and experiment on actually like walking around around the kitchen um opening drawers or um cabinets um moving dishes up into them for example it's another use case here

00:04:42 uh that is actually what we did in this bachelor thesis uh where we explored like how how could you not use necessarily Thiago but similar robots because Thiago is pretty pretty pretty weak um but to assist technicians in maintenance of trucks uh for example when you're switching switching tires and stuff like that I just really love working in unity

00:05:08 is the game engine we're using to actually create the digital twin I I want Unity to be almost like a simulator that that's my big hopes and also the human robot collaboration I want to actually participate in like creating a really good system for actually hosting in that kind of environment maybe you you hold something you hold it up to

00:05:33 Thiago and Thiago reaches up and grabs it from your hand stuff like that really really fun