A multi-device multi-tasks management and orchestration architecture for the design of enterprise IoT applications

Abstract

Enterprise Internet of Things (EIoT) is the next advancement in technology. EIoT allows the involvement of embedded devices to participate in business processes to automate enterprise operations. EIoT demands the modelling of services based on tasks to automate business processes and to allow easy adaptation to new business models. It is anticipated to enhance efficiency, align physical operations on a real-time basis and provide analytics for fostering more customer-centric business. The architecture is the cornerstone for any application, and currently, there exist a variety of architectures for enterprise systems and IoT addressing the requirements of the respective domains. The reference architectures of conventional IoT platforms perform orchestration at the service level, which, despite its success, does not align well with the demands of enterprises processes. This paper proposes a reference architecture to consider both the requirements of EIoT as well as the conventional Internet of Things (IoT). This paper is a novel attempt to propose orchestration at the task level, considering the business process modelling of enterprise systems. This level of orchestration is scalable and flexible due to the automation at a more granular level. In contrast, the service-level orchestration works well for general IoT application but is not suited for enterprises whose sole aim is the modelling of enterprise operation at the very task level. Therefore, a reference architecture named a multi-device multitasks management and orchestration (MDMT-MOA) based on a top-down methodology is proposed addressing the needs of IoT and enterprise applications. A case study is implemented on top of the architecture, and the performance is evaluated using various metrics related to EIoT. Results indicate that the proposed architecture is efficient due to lightweight payload, scalable, fault-tolerant and flexible to adapt with new business models. (C) 2019 Published by Elsevier B.V.