 So today I would like to discuss indeed my research on my PhD, so can you learn a second language from a robot Twitter. And you can see the robot that I used on the left, that's the now robot, and a very young child next to the robot on this slide. But first before I want to start with my talk, I would like to go to the year 2030. So that's 10 years, almost 10 years in the future. And I want you to imagine the classroom of the future. So typically when you think about teaching and education, you see a teacher in front of a classroom and you see children inside the classroom. The teacher is teaching the children in front of them, in front of the teacher. However, when I look at the future, I see that there's a robot next to the teacher that can help feature during the day when there are problems. Because if we look at classrooms in day to day life and also over the past three years, we see that there are many, many, many children entering the classrooms and their teachers in front of a larger classroom, classrooms are getting bigger. And the word pressure is also getting higher for the teachers. Because these children are there in the room, there are so many children, they all are different, they have individual differences. There are many children who have learning difficulties, their children ahead of other children, children come from different countries, they don't know the language yet, there are all these typical differences between all the children. And when there's only one teacher in front of the classroom, it's difficult to cope with all these difficulties with all these problems. So my suggestion is that the teacher can point at a certain child and say, well, you can go and play with the robots, practice the material that we just taught. And then after you practice it with the robots, you can come back to the classroom and then continue our lesson. And in that way, all the children can practice on their own with their own problems and they get some kind of individualized content. My name is Mirjam De Haas, as many already said, and I work in the El Tutor project. And the El Tutor project was a project where we taught children a second language using the now robots. And it was a European project with six universities and two companies. One of the companies also created the robots many years back. And yeah, we developed many, many, many experiments and a program for children to work with the robots. And in the Netherlands, we focused on two particular target groups. We focused on Dutch children learning English and immigrant children learning Dutch. And so if you look at that large classroom, for example, the world is becoming more international. And people are moving around to different countries and they take their children with them, their family with them. And then there might be children in the classroom who don't speak the language yet, the academic language yet. So then the robots can help these immigrant children learn Dutch. And in the same way, because the world becomes more international, English becomes more important. And it's a good idea to start early at an early age with teaching children English, especially because when you learn languages at an early age, you will become more proficient in the language. And in this presentation, I will focus on Dutch children learning English. So we already know from many, many studies that children can learn from robots. Their robots can teach children many stuff and they're quite good at it. As you can see in a review where you basically saw that robots are almost as good as human tutors. Of course, these are published papers. So let's take this with a bit of salt. But the experiments were of course also designed particularly for the robots to be as good as possible. But we did see that children can learn from robots already from the past three years. However, we're now at a stage at a point in the research that there isn't these for longer term studies. There should be more studies that have repeated interactions for children over time. Learn together with the robot and we can see how this interaction develops over time. Because when you show children a robot, and maybe that's even the case, or that is even the case with adults as well, that if you show that robot for the first time, the interest peaks. So there's a large novelty effect where people are interested in the robot. The engagement goes up. People are very, very interested in the robot. But over time that decreases because if the robot does the same thing all the time again, then yeah, people are less engaged with the robot over time. So it is very important to look at the robot during a longer time, longer periods. So that's what we did during the El Tutor project. We finished the project with a large scale experiment with many, many children. And all children had seven lessons. So that's a one month experiment where children saw the robot two times a week. And they were practicing different English words together with the robot. And also tableting from of them displaying the content, the environment of the words that they were learning. So for example, dragging more, so they had to drag animals into cages in the top right image. But also the words like sliding on a playground or throwing a ball to those kind of words children learned. And something that we were particularly interested in is the comparison of a robot versus a tablet screen, because children, they know how to work with screens. They know how to use a phone. They know how to use a tablet or a laptop or anything. But the advantage of having a robot there is that it is a robot is a physical entity. So there is something in the room that knows what you're talking about. And it has these arms and legs and hands, whereas can display images or gestures with. So for example, the example on the screen is a word monkey. The robot can actually use his arms and his hands, his body to express the words to depict the meaning of the word monkey with an iconic gesture. So we used that to test whether that actually works and whether children would learn more when the robots would use these gestures then versus a robot who did not use these gestures. This is another example that we used. So the word heavy and that the robot would express the word would use a gesture for the word heavy. What we found was, well, we have these two words and we have these gestures. However, some words are more iconic. You can express them in a more iconic way. So the word for for the gesture for monkey is very, very iconic. You can immediately see that it's a monkey when a robot does this gesture. But when the robot is using the gesture for heavy, it's a little bit less iconic because it can also be low, for example, or something else or high because it's also moving his hands up. So it's a little bit less iconic. And we also saw that in the results. We saw that children did learn more words with a robot using iconic gestures for a very iconic concept like monkey, but less, they learned less words when the robot was using an iconic gesture for a less iconic word such as heavy or father words. But what we did find was that children were more engaged with a robot, but there was a slight difference there. The children were more task-engaged with the robot without gestures. So they were more focused on the task on the tablet when the robot wasn't making gestures, but they were more engaged with the robot when the robot was using gestures, which kind of makes sense because if just a robot is standing next to you and it's just talking to you and not using any gestures, it makes sense that you're focusing on the task more. And when the robot is using fun gestures to look at, it makes sense that you have more energy next to the tablet. But what was interesting, interestingly, we did see that both of these engagements correlated with second language word knowledge. So children did learn in both situations more words when they were engaged either with the task or the robots than when they were not so engaged. So when we go back to the future, back to the future, yeah that's a strange sentence and familiar one, so when we go back to the year 2030 and we think of the robots in the classroom, so something is, so what I can say now is that we have to look into these gestures more, we have to look into the physical body of the robot and the embodiment of the robot, how it helps children in their learning more. So maybe using motivational gestures, for example, if iconic gestures don't work for the different topics, but then if we have that information, we can actually use that robot in a classroom to help individualized or to help children. Well, thank you.