High technology push for education and health
In a Swiss classroom, two children try to escape a twisted maze with the help of a rather cute little robot. Interaction is both simple and fun. It also provides researchers with valuable insights into children’s learning process and the conditions under which information is most effectively absorbed.
Rapid improvements in intuitive human-computer interactions (HMIs) are expected to bring major changes to society. In particular, two European research projects provide insight into how these trends may affect two key areas: education and health.
the child learns
ANIMATES, an EU-funded cross-border network of universities and industry partners, is investigating whether and how robots and artificial intelligence (AI) can help us learn more effectively. The concept revolves around making mistakes: children can learn by identifying and correcting the mistakes of others. And getting a bot to make those mistakes can be helpful.
Mr. Mohamed Chetouani, a professor at Paris Sorbonne University and coordinator of the project, said: “The professor cannot be wrong. “But a robot, yes. Mistakes are very useful from a pedagogical point of view.”
For Professor Chetouani, asking questions like “can robots help children learn better” is quite simple, because learning is a very complex concept. According to him, any automatic assumption that, for example, children who concentrate in class learn more is not necessarily true.
Therefore, the project focused from the beginning on asking more relevant and specific questions to help determine how robots could be useful in the classroom.
ANIMATAS consists of sub-projects, each led by an early-career researcher. One of the subprojects aimed to better understand children’s learning process and analyze the types of interactions that best help them retain information.
Robot roles
An experiment to answer this question invited children to team up with the robot QTRobot to find the best path on a map.
During the training, the robot communicates with the children and gives them advice and suggestions. It also carefully measures various indicators of children’s body language, such as eye contact and gaze direction, intonation and facial expression.
As they had hoped, the researchers confirmed that certain patterns of interaction corresponded with improved learning. With this information, they will be able to better assess the quality of children’s interactions with educational materials and develop strategies to optimize these interactions in the long term and therefore maximize their learning potential.
Next, the research team will try to discover how to adapt this enhanced learning in contact with the robot to children with special educational needs.
“We think it could be very important in this context,” Professor Chetouani said.
Help at hand
Aki Härmä, researcher at Philips Research Eindhoven, Netherlands, is convinced that robotics and artificial intelligence will revolutionize healthcare.
In the project PhilHumansJunior researchers from five EU-funded universities in Europe, which it coordinates, work with two commercial partners, R2M Solution in Spain and Philips Electronics in the Netherlands, to study how innovative technologies can improve human health.
Härmä says that artificial intelligence enables access to new services and “this means that healthcare can be delivered 24/7”.
He talks about the great potential that technology offers to help patients manage their health from home: apps that can monitor a person’s physical and mental state and detect problems early, chatbots that provide advice, diagnoses, and build algorithms that allow robots to move safely. a house.
Empathic Bots
The project, which started in 2019 and will end at the end of 2023, consists of eight sub-projects, each led by a doctoral student.
One subproject, led by Phillips researcher Rim Helaoui, is investigating how to encode specific skills of mental health professionals (such as empathy and open-ended questioning) into an AI-powered chatbot. Such an ability would allow people with mental illness to receive support while staying at home and therefore at a lower cost.
The team quickly realized that replicating the full range of psychotherapeutic skills in a chatbot created problems that could not be solved all at once. So he focused his efforts on one main challenge: building an empathetic bot.
“This is an important first step in getting people to open up and trust each other,” explains Ms. Helaoui.
To begin with, the team developed an algorithm capable of responding with appropriate intonation and delivering empathic content. The technology has not yet been developed into an application or product, but it provides a starting point that can be used in many different applications.
fast forward
PhilHumans is also exploring other ways to apply AI to healthcare. The algorithm, which is in development, uses “camera vision” to understand the tasks a person is trying to perform and analyze the environment.
The ultimate goal would be to use this algorithm in a home assistance robot to help people with cognitive problems perform everyday tasks.
Härmä says one of the factors driving the project forward is how quickly other organizations are developing natural language processors with impressive capabilities like OpenAI’s GPT-3. The project intends to take advantage of this surprisingly rapid progress in these and other areas, and move even faster.
ANIMATAS and PhilHumans projects are actively working to push the boundaries of intuitive HMI.
Thus, they gave young researchers the opportunity to gain valuable knowledge and introduce themselves to companies in the sector. Thanks to these two projects, a new generation of highly qualified young researchers is perfectly equipped to develop HMI and its potential applications.
Research for this article was funded through the EU’s Marie Skłodowska-Curie Actions (MSCA). This article was originally publishedThe horizonthe EU magazine dedicated to research and innovation.
To find out more about the EU-funded projects mentioned in this article, visit the links below.
ANIMATES
PhilHumans
