How to Change a Light Bulb in Your Smart Home: A Digital-Twin Based Approach.

The physical manifestation and the digital counterpart of smart devices are conventionally modelled as inseparable entities. Activity prediction models and automation struggle to comprehend occasional changes in a smart home. We demonstrate, in a simulated environment, how smart home modifications negatively impair the functionality of a smart home. Previously, we introduced a novel approach, Longevity Digital Twins (LDTs), with the aim to decouple the physical manifestation from its digital counterpart. This decoupling is inherently useful to allow the expression of relationships on the digital side while the physical manifestation can be changed with little configuration overhead. We build atop this approach to enable seamless management of the modifications in a smart home, such as adding, removing and replacing smart devices, whilst preserving learned context and automated behaviour. This is achieved by abstracting the physical environment into a hierarchical digital architecture composed of LDTs and a new class of digital twins denoted as Composed Digital Twins (CDTs) with the responsibility to build and maintain over time a holistic view on the deployment plane and the behavioural plane in terms of connections, interactions, and relationships.