Adaptive access to robotic learning experiences in a remote laboratory setting

Abstract

We describe the architecture of a small remote laboratory made available by means of a web-based adaptive system (Mindlab2). The system supports hands-on e-learning experiences with robotics and automation, performed via internet. It allows the learners to access and solve “problems”, submit their solutions and have them evaluated. The problems are defined by the teacher, also specifying instructional aspects, such as the learner's skills involved in producing a solution. For each individual learner a personal Student Model is maintained, recording learner's knowledge and skills, as shown by the solved problems. The access to the problem is guarded by the system, ensuring that the problems can all be “seen” by any student, whereas a learner can work on a problem only if its prerequisites are covered, or anyway “not too far” from her current competences. The available automata are modular structures of various origin (a robotic arm connected to an Arduino platform, and some Lego MindStorm builds, in various configurations). Such hardware is reasonably sophisticated in terms of functions, enjoys a wide provision of sensors and actuators, so allowing for presenting the learners with a good deal of problems, and is sufficiently affordable in terms of cost. Making the learner choose and sequence the problems to undertake, with the only limitation of the problem affordability measured by the student model, allows learners to follow a highly personalizable learning path in a pedagogically controlled setting, which we submit is conducive to better engagement and learning results.