For more than 15 years, EPFL researchers have been exploring ways to improve training for apprentices. The findings are summarized in a recently published book and a website for teachers, and new educational technologies are currently being developed.
Switzerland is known for its fondue, its punctual trains and its watchmaking industry. But the country has another claim to fame: its vocational education and training (VET) system. At the end of compulsory schooling, apprentices join a three- or four-year program that combines on-the-job training in a host company with classes at a vocational school. “The Swiss official discourse places VET somewhere close to Roger Federe” quips EPFL professor Pierre Dillenbourg, who spent 15 years as head of the Leading House DUAL-T, in the recently published book Educational Technologies for Vocational Training: Experiences as Digital Clay.
The book summarizes 15 years of research into digital technologies for vocational training, conducted by EPFL in partnership with the Swiss Federal University for Vocational Education and Training (SFUVET) in Lugano and the University of Fribourg. The number one priority is to build bridges between school and the world of work. “The VET system operates in silos,” says Dillenbourg. “What apprentices learn at school does not align with what they do in the workplace. That’s why we developed the ‘Erfahrraum’ model, which uses digital technology to bring the work experience into the classroom.”
Deriving value from technology
As part of this bridge-building work, the researchers have developed Realto, an online learning platform where apprentices can download and annotate photos and videos then share them with teachers and supervisors. They’ve also designed virtual reality tools that let users perform tasks that would be impossible in the real world, and gain insight into phenomena invisible to the naked eye. One example is the Tinkerlamp, a combined projector and small-scale model system for warehouse optimization. Another is the Static AR app, which allows apprentices to see and understand the forces acting on a roof structure.
The VET system operates in silos. What apprentices learn at school does not align with what they do in the workplace. That’s why we developed the ‘Erfahrraum’ model, which uses digital technology to bring the work experience into the classroom.
“Using technology to simulate reality isn’t especially interesting in and of itself,” explains Dillenbourg. “We wanted to explore how we could derive educational value from these tools. That said, technology isn’t a silver bullet. How helpful it is depends on how you use it.” A new website, eduscenarios.ch, builds on the work done by the DUAL-T team. It features a selection of technology-enabled learning activities that VET teachers can use to close the gap between the classroom and the workplace.
“We’ve developed 14 scenarios, all of which are easy to reproduce and adapt,” says Richard Lee Davis. The postdoctoral researcher at EPFL’s Computer-Human Interaction in Learning and Instruction (CHILI) Laboratory oversaw the design of the website and the scenarios. “We were determined to make these activities accessible to any teacher who wants to use them.” For each scenario, the website provides an overview, a description of the teaching approach, a “recipe” for using it in the classroom, a list of digital tools to support the activity, and a “tips and tricks” section.
EPFL is pursuing further VET-focused research through the Digital Vocation, Education and Training (D-VET) Hub, which is led by Prof. Tanja Käser, the head of the Machine Learning for Education (ML4ED) laboratory. For instance, the D-VET Hub team has developed interactive simulations for the classroom-based exploration of complex workplace scenarios: ChemLab is a virtual lab where apprentices can experiment with modifiable components; HeatingSim simulates a solar heating system with adaptative settings; and PharmaSim is a virtual pharmacy in which apprentices have to advise a customer.
“Our aim is to design innovative tools that can be deployed at scale in Swiss schools,” says Thiemo Wambsganss, postdoctoral researcher at the ML4ED lab. “We’re building these tools with input from a large number of teachers, and we’re using machine-learning methods to study how they affect learning and how they can be improved.” Wambsganss has developed WritingTutor, a customizable chatbot that analyzes the argumentative quality of a piece of writing and gives individual feedback. Research has demonstrated that the tool has shown a positive effect on the learning experience.
Our aim is to design innovative tools that can be deployed at scale in Swiss schools. We’re building these tools with input from a large number of teachers, and we’re using machine-learning methods to study how they affect learning and how they can be improved.
Wambsganss was also involved in designing RELEX (Recipe Learning through Examples), a system that apprentices can use to write recipes and get personal feedback on their performance. “A study involving 200 participants showed that adaptive feedback improves procedural writing and enhances the user experience,” he explains. The team’s next challenge is to persuade more VET teachers of the potential benefits of digital learning environments. “Our role is to innovate and to demonstrate what technology can do,” says Dillenbourg. “In more than 15 years of research, I’ve seen that educators are increasingly willing to embrace the possibilities of digital tools – and that our work has helped shape federal vocational education and training policy.”