A context-aware smart product-service system development approach and application case

Abstract

With advancements in information and communication technologies (ICTs), one of the epitomes of ICTs are smart product-service systems (SPSSs), which have attracted much research attention in recent years. The capture and analysis of users' dynamic situations are crucial for generating personalized services. Thus the context awareness ability with better self-adaptation and capability to mine information on implicit needs appears to be promising to cope with this issue. However, approaches for the development of a model that integrates context analysis and SPSS have not been fully explored. Therefore, this study presents a new perspective with the incorporation of context awareness, and then proposes a conceptual framework for an SPSS with context awareness (CA-SPSS) and the approach to develop this system. Specifically, the approach consists of four steps. The first step is context acquisition to create a wireless sensor network with wearable devices to acquire the physiological data of users. Second, context modeling procedures are set up on a cloud platform, such as for data collection, storage, and analysis. Moreover, machine learning datasets with information obtained through multiple channels, such as user, physical, sensed and inferred contexts, are generated. Afterwards, neural network models are used for context reasoning to determine the modalities for the users. To produce the appropriate service contents for each user modality level, a multi-criteria decision making process is employed to form customized service strategies for context distribution. To implement the proposed approach, an SPSS case for supporting non-professional sports competitions is developed and assessed in terms of the model performance and user satisfaction. The results show that the model deployed in the system has a better performance than traditional machine learning models. According to the findings from the user experiments, the subjects are very satisfied with the personalized service bundles and intelligence, but find that the reliability is suffering. The case study shows that the proposed CA-SPSS approach contributes to extending the available definitions of context in the literature and effectively guides the further development of SPSSs, especially in sports and mobile health.