MENU
Aim higher

Show some emotion to learn better

Europe needs more science graduates. But how to engage young people in scientific subjects? Can we ensure that initiatives such as gamified educational experiences are effective? And, if we can, how do we know that students are actually learning in virtual classrooms? The EmotionLab created by Danish and Swiss partners has the answers.

The modern world is based on scientific and technological innovation. But some two-thirds of science students in Europe drop out of their studies in high school or university. What can we do to engage and retain more young people in science and technology education?

One idea is to transform and enlarge the learning environment using technology and creativity developed in interactive virtual learning environments that use gaming elements to drive engagement and understanding. The EmotionLab project is creating new technology to detect emotions and apply this understanding to create cutting-edge adaptive learning solutions to enhance science education and corporate laboratory training.

Working on emotions

“The idea is to build the next generation of education tools,” says Michael Bodekaer, CEO & “Experience Researcher” at Danish company Labster. “We are taking insights and learning from the gaming industry to provide new ways to learn science in 3D learning simulations including science practical classes.”

Labster has created a virtual educational environment – the Learning Journey Island – that uses real life stories to bring subjects to life. “Using gamified stories and interactive visualisations of concepts only possible in the virtual world improves student engagement,” says Michael. “But it is important that we can prove this and show clear learning outcomes.”

This is where EmotionLab comes in. All students progress at their own pace and one key feature of virtual learning environments is that the course of study can be adapted to each individual student providing a personalised approach to education. But the essential element here is to find some way to measure student engagement remotely and then adapt the learning programme according to their mood or emotional state.

Poker face

Past approaches to establishing the emotional response of students have included brain wave monitoring using an electroencephalogram (EEG). EmotionLab’s approach has focused on creating a 3D image of each student’s face and mapping muscle tension. This approach means no need for EEG and its associated invasive paraphernalia and means the technology can be implemented anywhere in the world via, for example, a laptop camera.

Project partners Learn Technologies based in Switzerland are world leaders in applying biometric testing to education technology applications.

“By mapping muscle tension in a range of students and using machine learning to recognise ‘micro emotions’, we can reliably correlate the emotional state of a student,” Michael states. “The cues are subtle, but deep machine learning is fantastic for pattern recognition and as every poker player knows we all have a ‘tell sign’!”

The project has allowed Labster to refine its student algorithms and better identify when students are flagging or when they get that ‘Eureka moment’ and finally understand the material.

Pre-commercialisation is now underway with further testing happening with real student courses at Copenhagen University. Several papers on the research are also close to publication.

The breakthrough for widespread use of the technology will be increasing use of cameras capable of better depth sensing in laptops and smartphones, which is happening now.

The main initial market for this technology will be high schools and universities, but the potential market for immersive, virtual learning is immense. Large scale demonstrations of the technology involving simulations of otherwise hazardous laboratory scenarios, for example working with salmonella bacteria, and a full chemistry degree programme are planned.