The concept of Smart Learning Environment (SLE) has recently emerged with the aim of transforming current technology enhanced learning environments so that they can provide learners with adequate support at the right time and place based on their needs, which are determined by analyzing their learning behaviors, performance and contexts. SLEs are promising but also challenging especially in the case of physically-situated scenarios where participants interact with multiple devices, such as those proposed in Ubiquitous Learning environments based on the Internet of Things, or in the case of scenarios with a massive number of participants which are usual, for example, in Massive Open Online Courses.

There are two problems that hamper the success of SLEs in both types of scenarios. First, the design and redesign of increasingly effective learning situations are currently not informed by indicators of the impact in learning of previous design realizations. Second, the orchestration of learning situations is a daunting task for teachers and learners, that involves the monitoring, awareness, (self-)regulation and assessment of learning activities. Both problems stem from the fact that obtaining the adequate information required to make decisions about the (re)design and orchestration of non-trivial learning situations is out of reach for teachers and learners, given the high number of participants or the diversity of devices that can be involved in the scenarios.

Learning analytics can be considered a suitable approach to tackle both problems as they deal with the analysis of data about learning with the aim of understanding and optimizing learning and the environments in which it occurs. In fact, the potential of learning analytics to improve the support of teachers and students in different settings has already been shown. However, according to a recent report of the European Commission’s JRC, much research still needs to be done to tailor learning analytics for specific needs and contexts such as the aforementioned problems in SLEs.

The goal of this project is to improve the support of (re)design and orchestration of physically-situated scenarios based on different devices and massive scenarios within the context of SLEs by means of learning analytics. To do so, the project will propose (1) a set of learning analytics services that will provide indicators furnishing actionable information about the (re)design and orchestration, adequate visualizations of the indicators that will help participants make informed decisions that improve the (re)design or orchestration, and interventions that can be automatically triggered based on the indicators also to ameliorate them; (2) a framework for the integration of the proposed services in different SLEs; and (3) a set of pilot experiences showing how such services enhance the (re)design and orchestration.

To achieve this objective, significant challenges must be addressed: involve all stakeholders in the identification of indicators, visualizations and meaningful interventions; explore different data gathering and analysis approaches that can cope with very large datasets generated in settings with many learners and devices; address technical and semantic interoperability to deal with data from quite diverse sources; work on capacity building so that teachers can leverage the results of the project; and comply with data privacy and ethical issues.