Follow Us:

R-learning = improved learning results?

22.03.2013

What is R-learning?

In recent years, R-learning (robot-aided learning) has emerged as one of the efforts to adopt innovative technology in education for the improvement of education. By their narrow definition, R-learning refers to the “supply of young children education robots and contents according to the curriculum”. More broadly, it is defined as the “shift to digital education and the evolution of consilience integrating robot, contents, education program, and physical environment altogether” (KIST, 2013).

R-learning can differ depending on the type and role of the robot as well as its target group and applications offered: robots can work autonomously or be tele-operated, as (peer) tutors or tutoring assistants; they can be used by teachers, children and parents, for different subjects and offer conversation, edutainment and other services (Han, 2010). Unlike general computers and e-learning, robots can suggest the learners to start and response to them through autonomous recognition. Furthermore, they are able to interact with learners through direct contact, e.g. by hugging or petting, and provide physical activities to reinforce learning. Robots as peer-tutors have become the most dominant form of R-learning, followed by teaching assistant robots (Han, 2010).

The case of the Republic of Korea: the R-learning System Project

One example of R-learning through learning and teaching assistant robots can be found in the Republic of Korea. Since January 2010, the Korean Ministry of Education, Science and Technology (MEST) has been in the process of establishing a robot-based learning (R-learning) system as part of the Plan of the Advancement of Early Childhood Education as announced on the 9 November 2009. MEST’s R-learning project is designed to maximize the synergy of combining education and scientific technologies. Experts and practitioners involved in early childhood education expect that such robotic systems will play a role in helping teachers, thereby advancing the development of education systems and consequently contributing to promoting creativity and character development for young children.

In order to develop R-learning systems and foster their use for educational purposes the MEST has been working closely with the Korean Institute of Science and Technology (KIST). As part of the R-learning Project several measures have been taken, including teacher training, development and supplies of R-learning content materials as well as building an R-learning infrastructure and the development of the educational robots. For the first years of operation (2009-2010) 10% of all kindergartens in the Republic of Korea were chosen as pilot sites. Between 2011 and 2013 up to 50% of kindergartens are expected to be provided with robots (for more information, see KIST Leaflet).

R-learning in practice: Genibo edu and iRobiQ

The KIST R-learning System Project offers a clear example for how and in which educational contexts robots can be deployed. The intelligent robot dog Genibo edu is specifically designed for early childhood education, offering different “play-based learning” activities (songs, games, role plays etc.) to enhance children’s social and emotional development by interacting through emoticons, sound effects and motions (cf. KIST Leaflet).

Another example for use in primary or secondary education is iRobiQ: It works as a learning and teaching assistant, offering children’s songs, English learning contents and games as well as a communication tool that checks attendances, collects learner portfolios and gives parents insight into their children’s learning achievements. Among other administrative functions it also wirelessly connects to server computers that make it possible for teachers to create, up- and download learning and teaching materials.

R-learning = improved learning results?

Experts and practitioners involved in early childhood education related work expect that such robotic systems would be helpful for advancing the education system and promoting creativity and character development for young children. In terms of education system, R-learning enables integrated information management and education management and facilitates information sharing through creating communities among various stakeholders. Also, by supporting self-directed learning R-learning is expected to promote early childhood development.(cf. KIST Leaflet).    

However, these claims have not gone uncontested. Many concerns about the dysfunctional aspects of technology, over exposure to technology in early childhood have been raised among teachers and education sectors. To address these concerns, bunch of studies have been conducted for a long time, and the impact of R-learning on early childhood education has been assessed in the area of physical, linguistic, cognitive, social and emotional development. One of the interim reports, “The impact of education based on R-learning robot platform(R-learning) on children’s pro-social behavior and emotional expression” (2006), a study conducted by KIST R-learning Center collaborating with experts in Korea Early Childhood Education Society, tried to measure the impact of R-learning systems in the area of social and emotional development of children. The study was conducted with 50 children in a randomly chosen kindergarten. These homogenous children were divided into two groups and the children in the experimental group who were treated with R-learning system showed significant difference with the children in the control group. They showed (positive) progress in both pro-social behavior and emotional development, in line with many other precedent studies (Davis, et al., 2005; Dautenhahn, 1999; Koizima, 2002; Koizima, et al., 2005; Ruvolo, Fasel, & Movellan, 2008; Tanaka et al.,2006, 2007; Tanaka & Movellan, 2006; Werry et al., 2001; Werry & Dautenhahn,1999)(recitation)(KIST, 2006).

Even though a certain consensus has been achieved by communities involved in r-learning about the fact that educational robotics are valuable, the long term effects of R-learning are yet to be  measured and rigorous research methodologies are still an issue to secure validity and reliability of the evidences. Numerous workshops, conferences and publications have focused on the issue. Yet many of these show certain sub-optimal properties, such as referring to very restricted experimental populations, and the missing of a pedagogical and methodological framework as well as significant quantitative and qualitative evaluations (Bredenfeld, Hofman, & Steinbauer, 2010). The involvement of multiple (at times also competing) stakeholders, especially the diversity of technologies and approaches makes such a direct comparison and evaluation nearly impossible. As a result, a more systematic and validated foundation and methodology are needed so as to make R-learning a more valuable and reliable approach in education.

Many concerns are being raised on the subject of robots’ use in educational settings and a rigorous research needs to be done in order to make sure R-learning can be put into practice in a scientifically founded and pedagogically meaningful way. Despite it being only at an emerging stage for now, it already appears to have a high potential and will certainly encourage the development of new and rich ways of teaching and learning in the future.

References and further readings:

  • Bredenfeld, A., Hofmann, A., & Steinbauer, G. (2010). Robotics in Education Initiatives in Europe - Status, Shortcomings and Open Questions. In: Intl. Conf. on SIMULATION, MODELING and PROGRAMMING for AUTONOMOUS ROBOTS, pp. 568-574. Also accessible online via: www.terecop.eu/simpar2010/TR-TWR-2010/22-TeachingRobotics.pdf
  • Davis, M., Robins, B., Dautenhahn, K, Nehaniv, C. L., & Powell, S. (2005). A comparison of interactive and robotic systems in therapy and education for children with autism. In Proceeding of Assistive Technology from Virtuality to Reality, 8th European Conference for the Advancement of Assistive Technology in Europe (AAATE '05), Lille, France, 6-9 September 2005, pp. 353-357. Amsterdam, The Netherlands: IOS Press.
  • Han, J. et al. (2009). How different in cultural acceptance of tutoring robots serving augmented reality? In: International Conference On Advanced Communication Technology, pp. 2006-2008.
  • Han, J. (2010). Robot-Aided Learning and r-Learning Services, access online via:  cdn.intechopen.com/pdfs/8632/InTech-Robot_aided_learning_and_r_learning_services.pdf
  • Jung, J.-H. & Bang, Y.-S. (2011). A Study of the Use of R-learning Content in Kindergartens. In: 8th International Conference on Ubiquitous Robots and Ambient Intelligence (URAI), pp. 708-710.
  • Kanda, T., Sato, R., Saiwaki, N., & Ishiguro H.A (2007). Two-Month Field Trial in an Elementary School for Long-Term Human-Robot Interaction. In: IEEE Transactions on Robotics, pp. 962-971.
  • Kozima, H., Nakagawa, C., & Yasuda, Y.(2005). Interactive robots for communication-care: A case-study in autism therapy. In Proceeding of IEEE Workshop on Robots and Human Interactive Communication (RO-MAN 2005), Nashville, Tennessee, 13-15 August 2005, pp. 341-346. Piscataway, NY: IEEE Press.
  • KIST (2006). The impact of education based on R-learning robot platform(R-learning) on children’s pro-social behavior and emotional expression(Korean), pp.75-87
  • KIST (2013). Robot Based Learning for Early Childhood Education in Korea. Presented at the UNESCO Regional Consultation Workshop on ICT in Education Policy, Infrastructure, and ODA Status in Selected ASEAN Countries, Bangkok, 5-6 March 2013.
  • KIST website (engl): eng.kist.re.kr/kist_eng/main/
  • KIST Leaflet on R-learning (engl): r-learning.or.kr/new/leaflet/KIST_leaflet_EN.pdf
  • Ruvolo P., Fasel I., Movellan J. (2008). Tuning Optimizers for Time-Constrained Problems using Reinforcement Learning. Neural Information Processing Systems (NIPS) Workshop on Optimization for Machine Learning.
  • Tanaka, F., Cicourel, A., & Movellan, J.(2007). Socialization between toddlers and robots at an early childhood education center. PNAS, 104(46), 17954-17958.
  • Tanaka, F., & Movellan, J.(2006). A way of implementing an internal reward for autonomous robots. International Workshop on Synergistic Intelligence Dynamics at Humanoids. Genova, Italy.
  • Tanaka, F., Movellan, J. Fortenberry, B., & Aisaka, K.(2006). Daily RI evaluation at a classroom enviromnemt- Reports from dance interaction experiments. Proceedings of the 2006 ACM Conference on Human-Robot Interaction, 3-9, March 2-4, 2006, Salt Lake City, Utah, USA.