Personal and social robots in education28 November, 2018
The role of robots in the education of the future
Whether robots will replace the work done by humans on a large scale within the next 10 years is currently a subject of heated debate. One question, in particular, stands out from this roundtable: could teachers be replaced by robots?
According to Martin Hamilton  (Jisc’s  resident futurologist), the technology could boost a teacher’s potential both in schools and in higher education. Hamilton argues that a teacher who has 30 pupils or more in a classroom needs help and a robot assistant “never gets tired, never gets bored and can work in all subjects”. During the same interview, Professor Anthony Seldon (historian and Vice-Chancellor of the University of Buckingham) asserts that new technologies will make it easier to provide personalized education and give teacher-student ratios of 1:1 instead of 1:100 or 1:30, as at present. Seldon believes that this type of new technology (such as the introduction of robot assistants and artificial intelligence  in education) will help in developing a personalized study programme for each student and, in turn, students will benefit more from the classes, attaining attention levels close to 100% instead of the current levels of 20-25%. In his latest book  (Seldon, 2018), he assesses the future impact of AI in education. Up until now, the prediction that all jobs may be replaced by intelligent robots in the near future had not gained any major traction in those professions in which personal interaction is an important factor. However, this is starting to change. Education is already starting to feel the impact, with the use of data to simulate student behavior and develop intelligent tutoring systems.
An article published in MIT Technology Review  discusses the conclusions of the report published by Pearson, the British innovation foundation Nesta and future-gazers from the University of Oxford’s Oxford Martin School. This document studies what effects replacing jobs with robots will have and what people’s roles will be in this, defining what human skills will be in demand in 2030 . Five are highlighted in the United Kingdom , which can be easily extrapolated internationally: judgment and decision-making, fluency of ideas, active learning, learning strategies and originality. It suggests that skills such as creativity, adaptability and judgment will be more important than, for example, knowledge of a particular subject or use of a tool, as the former represent intrinsically human skills that are hard to find in a machine, while the latter can be carried out by robots and other systems that use AI.
The purpose of this article is to review the recent developments in robotics that can be used in learning and for performing educational tasks.
A domestic robot is a type of self-reliant service robot  that performs housework, thereby improving the quality of life of those who live in the house. It can also be used for education, entertainment or therapy. Some of these robots have a Wi-Fi connection to intelligent environments and are endowed with a high degree of self-operability thanks to the progress achieved in fog computing  and domotics.
There are different classes of domestic robots. Most of them are used to help in performing routine tasks and optimize the use of time, while others are used for entertainment. The classification can be simplified to indoor robots (for cleaning floors, ironing, cooking, security and surveillance), outdoor robots (cleaning gardens and yards, mowing the lawn, cleaning swimming pools), play robots (humanoid robots, electronic pets) and social robots (used to educate, provide company and assistance to people – such as the elderly and patients, and for telepresence).
However, the nomenclature that classifies certain robots as educational robots refers to those that are used as an educational resource and fulfill a teaching purpose. Their main goal is to develop a body of knowledge while fostering student engagement through emotion. Some of these could be included in some of the categories mentioned earlier. Robots have become an educational tool in many primary and secondary schools, increasing the interest in programming, AI and robotics. Universities are starting to include robot programming in their engineering degrees.
In our particular case, we will analyze the role played by the robots that are broadly classified as personal, social or educational robots within learning-related tasks.
-The role of personal robots in learning
In the domestic sphere, there is a type of robot called personal robots. As we have seen, they can perform a series of tasks that make daily life easier. In the particular aspect we are interested in, education, there are various models that help in the learning process and are intended for use in both child and adult education, in language learning or other disciplines.
Temi the personal robot provides insights into the role of robot-driven interactive learning. The video shows different functions typically performed by personal robots, among which we will highlight language learning for children.
Buddy is advertised as an “emotional companion robot”. Alongside surveillance, providing company, taking and making calls and other actions that are also performed by Temi, Buddy is also able to teach people basics of software programming.
Buddy, the emotional robot
-Robots in educational centers. Teacher robots and telepresence
Although there are already many different robotic devices that teach students to learn different subjects and tasks (from mechanical arms for engineers to simulators), we will focus on how robots are entering classrooms to teach a variety of subjects, from languages (native or foreign languages) to social skills. They can even provide reinforcement in teaching a particular subject or in attending to students, becoming in many cases teaching assistants or even actual teachers. In other cases, however, they become a sort of avatar for the students who opt for the telepresence system; although in these cases their role does not arise from AI but from the possibility of letting students interact remotely with their classmates.
The Pepper programmable robots and their “little brother” Nao (which we will get to later) have been developed by the company SoftBank Robotics. These two robots are currently evolving within the education sector as teacher assistance tools and also as platforms for learning how to program, from primary school to higher education.
Thanks to its appearances in the media, Pepper is a fairly well-known case in which several types of robot have been merged into a single machine. This humanoid robot has been endowed with emotional intelligence and there are secondary and higher education centers that are planning to use it for learning assistance. One case is Pepper’s presence in language classes at the Shoshi High School in Waseda, Fukushima prefecture Japan, where the robot studies together with the human pupils. Pepper is also gaining a foothold in European higher education. At the latest 4YFN (4 Years From Now), Pepper acted as assistant in the simulation of an international business class given by professor Pedro Parada from ESADE Business School . The intention was to show how the robot can animate and personalize classes, encouraging student participation by interacting with them, getting to know them personally, and answering questions about class content and planning. The robot will also enter the classrooms of the London Design and Engineering University Technical College to assist in teaching cutting-edge robotics, together with 12 units of the small Nao.
Pepper and NAO in the London Design & Engineering UTC
At the Paul Bert school in Poitiers, France, Nao teaches French grammar to children, first showing them a video and then giving a questionnaire that the pupils answer as a group. The robot has stored a series of resources in its database and uses audiovisuals to hold the pupils’ attention. The teacher, Serge Evreinoff, is with them, supervising the class and telling the pupils when it is their turn to speak. At the end, he gives his opinion about the questions asked by Nao, explaining that the pupils have been more participative than usual, with a greater degree of enthusiasm and attention, and unafraid of adults’ opinions or judgments. Evreinoff wonders what role this type of device could play in day-to-day school life. He believes that the “Nao of tomorrow” will support learning individualization: “The computer must become an extension of the teacher”. He also says that “robots can help children and teenagers acquire a greater awareness of computer programming”.
A frame of the video NAO at School in Poitiers, France, with Canopy Network (https://youtu.be/vzAfuWFt9cg).
At the University of Montana in the USA, instead of having a robot teacher, it is the students themselves who have become “robotized” . The offsite students in the classes interact with the onsite students through cameras on remote-controlled robots equipped with wheels and a screen. The hope is that this will enable students hundreds of miles away from each other to take part in the same classes and feel more connected with the group than would be possible using standard videoconference links.
Every day, before the students arrive for class, the teacher leading the Bill McCaw initiative turns on the Double Robotics robots, disconnects them from their charging stations and organizes them so that they are ready for the students. These robots are a kind of small Segway with an iPad on top. Meanwhile, the remote students log on and choose the robot they want to use (each one has a name, like Aristotle or Rosa Parks). The remote students can see their fellow students through the video screen and also control the device to move around the classroom or turn to face the other students and interact with them.
The PhD programme in Educational Psychology and Educational Technology (EPET) focuses on studying human learning and development and on different technologies that support learning and teaching. During a course held in spring 2015 , the students took part using an Apple iPad placed on a robot that swivelled on a fixed pedestal; another student interacted via a robot that could move around the classroom. The revolving pedestal, called Kubi (made by Revolve Robotics), lets telepresence students be a part of university and school classes. A drawback? Without the wheels, Kubi needs someone to move it around and put it in the classroom. However, the same Ph.D. programme has now started to use another type of robot, the Beam Robotic Telepresence made by Suitable Tech. They are similar to Double Robotics’ mobile pedestals but have an integrated camera and screen.
Video from MSU (http://education.msu.edu/cepse/epet/default.asp)
Smart toy robots used in learning and accompanied learning
Some robots, which are more attractive more affordable for the general public, are marketed in a toy format, even though their purpose continues to be education  and children’s education in particular. Some of these robots can be programmed by the children and, in other cases, they play the role of companion to guide or help them in their learning process.
-Learning coding and programming
Different types of devices are available for learning robotics and programming. Some of them use mobile phones: with the phone, users can control the toy using an app. There are a number projects that are still in an incipient stage of development or which are raising capital through crowdfunding websites; one of them is Romo, a small robot that uses a smartphone for its brain. Using another mobile device and a multi-platform app, users can move it around, make it show animated facial expressions, make it dance or turn it into a spybot. Romo is simple and very different from the typical image of the personal robot. In order to find another, more complex type of robot (also funded by crowdfunding), we need to turn our gaze towards products such as Antbo, an insect-shaped robot designed to help children learn robotics and programming. It has a scanning system that enables it to perceive and explore its environment, using its neural system to process and react to stimuli. Antbo has an accelerometer and touch, sound, light and distance sensors.
A variety of robot kits are available to introduce children and teenagers to robot programming, using systems such as Raspberry Pi . One example is Lego Mindstorms, a hardware and software platform produced by Lego to develop programmable robots based on Lego building blocks. They have a smart computer that controls the system, a set of modular sensors and motors and other components to create the mechanical systems – everything you need for a kit to help you start learning to programme at home.
-Learning languages, school reinforcement, special education
The Musio robot has been designed for educational purposes. Its job is to help Japanese children learn English through playing and improve their pronunciation and vocabulary. Musio talks while expressing complex emotions, thus practicing a type of conversation that is close to that used by humans. According to the company that created it, Musio has unlimited conversation powers, with the ability to hold unprogrammed dialogues thanks to its AI-based technology. Musio is accompanied by learning materials and an ergonomic optic reader-type extension called Sophy to help children study.
Source Nico video news (http://news.nicovideo.jp/watch/nw3642379)
Video advertising Musio
Practice using Sophy
Leka is a smart multisensorial toy developed by therapists and researchers. It offers children with special needs the possibility of playing with entertaining educational toys that encourage social interaction, improve motor, cognitive and emotional skills, and stimulate independent learning. The company developing it is currently raising capital to launch industrial-scale production of Leka. However, Leka can be acquired now through its alpha programme.
–Entertainment and company. Storytellers
Luka is a children’s reading robot designed by Ling Technology to encourage children to start reading. The procedure is to put the open page of an illustrated children’s book in front of Luka. The robot recognizes the story from its cloud-based library and reads it out loud using AI. It is able to recognize hundreds of books available on the Chinese and English markets, although it works best with a book chosen from its own library.
Studies on learning accompanied by social robots. The case of MIT Media Lab’s Personal Robots Group
The Massachusetts Institute of Technology (MIT) Media Lab’s Personal Robots Group focuses on the development of principles, techniques and technologies for the development of personal robots. Dr. Cynthia Breazeal and her students are carrying out research to improve smart robots that interact socially with human beings, with the aim of boosting the social and intellectual benefits they can bring. The robots work with humans as companions, learning from people, and fostering a more attractive interaction with them. The group is researching the impact of long-term, personalized human-robot interaction (HRI) applied to quality of life, health, creativity, communication and education, performing experiments in human subjects, both in the laboratory and in real-world environments.
-Case studies: Parle, Solero and Tega
A number of studies have been performed on children’s interactions with robots and education. One study was is about curiosity, where they used the robot Parle, the curious robot.
This project studies how interacting with a robot can awaken children’s curiosity and desire to know and learn. Children are invited to play a tablet game with Parle. It is a personalized assistive robot that learns English and behaves like an inquisitive child, expressing enthusiasm about learning and exploring. The MIT’s studies came to the conclusion that Parle’s curiosity “is contagious”. The children who had interacted with Parle in the study showed more curiosity after the interaction than the children who had not. Parle personalizes the interaction with each child: it optimally evaluates the children’s reading skills and uses latest-generation automatic learning algorithms to create a personalized tutoring system.
Another project is Solero (Socially Learning Robot), a robot that helps students engage more with studying through recognition of facial expressions using a Kinect sensor and people’s responses, creating a kind of emotional bond with the robot to increase the desire to learn.
The last project we will review is Tega, a robot that has been designed to support interactions with early education applications, from learning to read and write to storytelling. Tega is an improvement on the so-called “DragonBot”. This robot tells stories to children and then has a conversation with them about the stories, checks that they understand the plot and vocabulary, and gives emotional or inferential indications about the characters’ feelings. It tests curiosity by asking what will happen next and adapts its suggestions and reactions to the child’s verbal and physical responses .
Source: Personal Robots Group (MIT Media Lab)
In a technological society in which robotics is moving forward at high speed, it is inevitable that education, as a sphere that is essential for the advancement of society, should feel the impact. As the article has shown, robotic devices have been gaining a growing presence in the home. However, long before that, they had already become widely used in the professional sphere, starting with industry. They have also made their mark in schools and universities, either as teaching assistants or as study material. Robotics is now being driven by AI; robots have ceased to be pre-programmed automats and are now entities that learn with us. We do not know what their role will be in the immediate future but no doubt we will have to get used to seeing physical effigies for the intangible beings that, until now, we had only known as assistants on our phones and computers. They seem destined to play a role as companions in the future and without doubt, we will learn a lot from them. Or perhaps it would be better to say both from them and from ourselves as we relate with them. Who knows? Perhaps the robots will help us become more human…
 Podcast “What the Edtech?! The rise of the robots #9. Is artificial intelligence the biggest shake-up for tech since the invention of the car?”, Education Technology, 2018. Available at https://edtechnology.co.uk/Article/what-the-edtech-the-rise-of-the-robots [Accessed 12 July 2018].
 An organization that offers digital solutions for education and research in the United Kingdom.
 AI in the rest of the text.
 Seldon, A., Abidoye, O. The 4th Education Revolution. The Fourth Education Revolution: Will Artificial Intelligence liberate or infantilize humanity?, The University of Buckingham Press, 2018.
 What Skills Will You Need to Be Employable in 2030?What Skills Will You Need to Be Employable in 2030?, MIT Technology Review, 2017. Available at https://www.technologyreview.com/the-download/608981/what-skills-will-you-need-to-be-employable-in-2030/ [Accessed 17 July 2018].
 Looking toward 2020, in this case, and the future of work, see also “The Future of Jobs”, World Economic Forum, 2016. Available at http://reports.weforum.org/future-of-jobs-2016/ [Last accessed 20/07/2018].
 The future of UK skills: employment in 2030″, Nesta, 2017. Available at http://data-viz.nesta.org.uk/future-skills/index.html [Accessed 17 July 2018].
 A robot that can operate with a certain degree of autonomous decision-making power.
 Fog computing is a decentralized computing infrastructure where data, computing, storage and applications are distributed in the most logical and efficient place between the data’s source and the cloud. The objective is to improve efficiency and reduce the quantity of data sent to the cloud for processing, analysis and storage. It can also be used for security reasons. Fog computing applications include smart networks, cities and buildings, vehicle networks and software-defined networks. Definition of “fog computing” taken from Cutting edge: IT’s guide to edge data centers Available at https://internetofthingsagenda.techtarget.com/definition/fog-computing-fogging [Accessed 12 July 2018].
 Riverola, C. (2018) El robot Pepper hace de asistente en una clase de universidad, EFE Futuro, [en línia] https://www.efefuturo.com/noticia/robot-pepper-asistente-clase/ (accessed 07/17/2018).
 Johnson, S., “Robot Students? College Classrooms Try Letting Far-Away Students Attend Via Remote-Control Stand-In”, EdSurge, 2017. Available at https://www.edsurge.com/news/2017-05-11-robot-students-college-classrooms-try-letting-far-away-students-attend-via-remote-control-stand-in [Accessed 18 July 2018].
 Schaffhauser, D., “Robotics Enter Hybrid Instruction”, Campus Technology, 2015. Available at https://campustechnology.com/articles/2015/09/01/robotics-enter-hybrid-instruction.aspx
[Last accessed 25/07/2018].
 Recommended academic paper: Castro, E., Cecchi, F., Valente, M., Buselli, E., Salvini, P., Dario, P. (2018) Can educational robotics introduce young children to robotics and how can we measure it? Journal Comput Assist Learn; 34:970– 977 https://onlinelibrary.wiley.com/doi/epdf/10.1111/jcal.12304
 The projects started by Raspberry Pi by can be found at https://projects.raspberrypi.org/en/. One of them is to design your own robot using CSS programming (https://projects.raspberrypi.org/en/projects/build-a-robot).
 (2017) The New Robot in School. Tega, the fuzzy friend who tells stories to kids, MIT Spectrum [online]. Available at http://spectrum.mit.edu/fall-2017/the-new-robot-in-school/ [Last accessed 03/09/2018].
*Featured header image source: Kombusto. MIT Media Lab/Personal Robots Group.