Abstract:

The digital twin paradigm is based on representing objects from the real world (physical objects) in the virtual world (logical objects). Originating from the manufacturing domain, in which a part can be tracked during its entire production lifecycle and then during the entire exploitation phase, the digital twin concept has been used successfully in other domains. For example, in healthcare, a digital patient can give a holistic, multi-faceted and longitudinal representation of a patient, including her/his health state, medical history, genetic predispositions but also cross-cutting concerns such as lifestyle. Using methods from multi-agent system research, simulations can be made to predict future risks, the onset of chronic diseases. Another example comes from the cultural domain, in which artifacts from museums can be accessed through their digital counterpart, which can be queried for information or can be related with other digital counterparts, to realize more attractive and culturally rich interaction scenarios. From these examples we can highlight some of the features of digital twins: monitoring of a physical object, providing more services through the virtual interface, ability to form associations with other digital twins, prediction through simulation.

In this paper we are studying the feasibility of using digital twins in the education domain, specifically in the management of assets at university level, with a focus on managing research and didactic laboratories. Asset management is a critical administrative task, with effects on research and education quality. One of the main weaknesses in universities is the sub utilization of well-equipped laboratories, due to security reasons but also due to scheduling constraints. In research, common laboratories can be used by multiple teams, and also it is common in grant writing to specify the available equipment that can be used for grant realization. In addition, labs are frequently made available to the community, for example maker labs with 3D printers, that can be used by local artisans or entrepreneurs.

Existing platforms for asset management are static, can at most be used for reservation or for reporting, and are difficult to be extended with additional functionalities as they are part of larger, enterprise-level software systems.

We provide a novel approach, in which laboratories, as well as existing equipment, are represented as logical objects, with intelligent virtual interfaces, exposing physical laboratories` capabilities through decoupled, discoverable, semantic services that can be composed for implementing complex workflow scenarios. The functional and non-functional requirements documented in the paper describe on achieving new, seamless capabilities, for example automatic matching between research proposals or activities and assets, including the automatic composability of assets for more complex research tasks, semi-automated maintenance and parts ordering, security and auditing. Finally, a general software architecture is described, based on micro-services and virtualization, based on the functional and non-functional requirements.

Keywords:

e-learning, digital twin, asset management.