Automated IoT-Based Performance Assessments Through Activity Recognition and Semantic Evaluation in Smart Learning Environments

In this paper, we utilize internet-of-things and semantic technologies to automate the assessment of socio-technical behaviors in engineering students in a circuits laboratory setting. Socio-technical learning objectives in engineering courses, such as the ability to make ethically informed judgments and the proactive observance of safety procedures, are challenging to assess for many reasons, among them being that their assessment can be time consuming to administer and can be subjective. In this work, we propose a novel framework that includes perception and data aggregation layers that utilize both sensor and video monitoring approaches to perform activity recognition, an ontology schema for representing and relating the data, and a semantic query layer that reasons out the desired performance evaluation. Finally, our work presents conclusive data supporting the advantages of using IoT and semantic technologies in automating the assessment of socio-technical learning outcomes in engineering courses.