Abstract:

E-learning is now widely used at university level to replace traditional lectures and exercices by digital document reading, quizzes, tutorials, videos, demos and simulations. However the most challenging part is to replace hands-on labs by virtual or remote labs. The difficulty is not only tied to technology (real-time synchronism, scheduling 24/24-7/7, security, mobility, interaction) but also to psychology and cognition aspects. Indeed in distance learning, there is no physical presence of a teacher or tutor to supervise students and drive their motivation, so that the student-teaching process becomes an open loop. Therefore there is a need to find a substitute for the teacher, thanks to technology help, in order to make it a closed-loop process, that is regulated even if there is no teacher in the loop. One needs to be able to evaluate the student mood (concentrated, bored, amused, perplexed...) and to boost the motivation in learning with appropriate tools (immersion, gaming, feedback).

Here we address the case of a remote lab in electronics for undergraduate students and we evaluate both the motivation and satisfaction of students experiencing this distant practical activity.

We have developed a platform called LaboREM (for Remote Laboratory) that is in use since 2011. It consists of a hardware setup accessible through the web (server with instruments and Device Under Test DUT).

In addition to the electronic setup, a robotic arm and a mobile camera are used to mimic the hand and the eye of the student that might see the lab and act on it by choosing components to build the DUT he wants to characterize.

The lab architecture was presented at ICERI'15 and the use of computer vision to enhance the lab was proposed in Edulearn16. A teaser illustrating the concept is available in French: www.youtube.com/watch?v=98b1j5HIqn0.

On the client side, to infer the state of mind of the distant student, a webcam captures the video of his face. Image processing and pattern recognition are used to analyse face expressions and to detect motion (head, eyebrows, mouth etc). Depending on the captured information, various actions may be proposed (eg. make a pause, be quite, check something) and signals may be sent to alert the student. This allows to control the student behavior and compare it with a normal state.

On the server side, two cameras are used: a wide-angle Pan-Tilt-Zoom camera that is static, and a micro-camera onboard a quadcopter-minidrone that can fly in the lab on student's request, to see precisely a specific part of the equipment (eg. front panel of an oscilloscope). This increases immersion of the student in the remote lab and makes him the actor of the activity (instead of remaining passive).
As regards the pedagogical strategy, a game-like approach is implemented: it is based both on the use of a Top 10 (best anserws to questions during the measurement real-time activity) and on the use of a track-game to find the right solution to an unknown circuit. This is inspired by e-games scenario (levels, lives, points, timer).

Before and after the learning activity, the students answer questionnaires about satisfaction, motivation and psychological profile. From those questionnaires and from the scores obtained in the lab, we analyse the impact of game, robot, drone and camera video processing on the teaching process. We show that the behavior of students is highly affected by these tools, in a positive way as regards motivation.

Keywords:

e-learning, electronics, face analysis, game-based learning, labwork, mobile camera, pattern recognition, quadcopter drone, remote lab, student motivation.