Skip to content

The Future of Schools: Adaptive Learning, Robo-Teachers and More

Last week I spend two days in Vienna with Accenture since I’m part of their Female Talent Program including regular meetings and workshops. This time’s motto “Don’t think out of the box. Think like there is no box” turned out to be a two-day hands-on design thinking workshop on the future of education.

Design Thinking

Design Thinking is a methodology that puts the user at the center of the design process to solve complex or ill-defined problems. The process is iterative and can be understood as a collaboration between designers and users.

Learn more about Design Thinking on The Interaction Design Foundation

After conducting interviews and getting to the core of the problem of current education such as
– “not enough time for students”,
– “lack of personalization”,
– “every child has different needs”, …
we started to generate as many ideas as possible. We refined them, broke them down and built them up again and rated them before we started to prototype our idea. During a presentation at the Wiener Kammeroper we could test our first concept by gathering feedback by the audience.

Curious what we came up with? Well, my team was working on an adaptive learning companion that can assist children in learning by personalizing the content, interaction form, pace and difficulty to the needs of the child. Let’s break this down into manageable pieces:

Adaptive Learning

In general, Adaptive systems modify the system’s parameters based on the characteristics and needs of the human user without instructions by the user but rather by gathering data on performance, behavior and physiological measurements of the user.

Adaptive systems come handy in education. One example is Intelligent Tutoring Systems: ITS are computer-based, adaptive learning systems that customize the content of the task and the interaction style to the student’s needs in real-time.

One might think that even the most adaptive learning system can’t be as good as a human teacher. How can a machine take into consideration the affective state of the student and act accordingly? How can a machine use different communication strategies?

Researchers of the University of Iowa were able to train an adaptive learning system to adjust the used communication strategy based on the emotional state of the students. Moreover, the scientists could demonstrate that the intelligent tutoring system was able to increase motivation, confidence, satisfaction and performance of the students. All of them are clear indicators of learner’s success.


As a team we wanted to use the benefits of adaptive learning systems and combine them with some form of embodiment for our learning companion. We were thinking of a hologram or robot as a learning companion.

The representation of a learning companion in a tangible and/or visible form brings several advantages, especially when working with children. It’s easy and fun to interact with and the companion and the child share the same physical world. Next to that, we want to facilitate social bonding between the student and the learning companion. To sum up, learning with an embodied companion acting as a peer and not as a teacher should be fun for the children and therefore increase their learning motivation and outcomes.

Researchers at MIT Media Lab created a robot for language development of preschool kids. This robot was able to adapt his behavior to the language level of the child.

MIT’s Dragonbot Green with a preschool child. Learn more here.

They found out that kids learned more words, told longer, more elaborate and more complex stories when learning with an adaptive social robot as a learning companion. Next to that, most of the children reported that they consider the robot as a friend and that they enjoyed playing with him.

To sum up, I had two amazing days working on the future of education. The solution proposed by our team was derived from user insights and through empathizing with students. Next to that, we could find empirical evidence from the Human Factors field proving the core components of our idea.

Yang, E., & Dorneich, M. C. (2018). Affect-Aware Adaptive Tutoring Based on Human–Automation Etiquette Strategies. Human Factors, 60(4), 510–526.

Kory Westlund, J., & Breazeal, C. (2015). The interplay of robot language level with children’s language learning during storytelling. In Adams, J. A., & Smart, W. (Eds.), Proceedings of the 2015 ACM/IEEE international conference on Human-Robot Interaction (Extended Abstracts). ACM: New York, NY.