Decentralising IoT-Aware Business Processes Balancing Business Risk and Energy Consumption

As Internet of Things (IoT) devices become more capable, they participate in business processes (BPs), performing process activities and making control decisions that were traditionally assigned to central systems. This creates a design-time decentralisation problem: deciding where each process element should execute given limited device resources, communication failures, and battery constraints. Existing decentralisation approaches for IoT-aware BPs mainly focus on operational criteria, such as communication overhead, energy consumption, latency, or reliability. However, they provide limited support for analysing trade-offs between competing criteria and typically do not consider the business impact of failures. This paper proposes DECR, a decision-support methodology for decentralising IoT-aware BPs by jointly considering energy consumption and business risk. DECR models risk as expected business loss by combining reliability and business value, and formulates the decentralisation problem as a bi-objective generalised assignment problem, which is solved using an adapted bi-objective simulated annealing metaheuristic. Rather than a single solution, DECR provides process stakeholders with potentially Pareto-efficient alternatives to support the selection of decentralised models according to energy constraints and risk appetite, enabling business-aligned decisions. The methodology is illustrated through an irrigation case study based on a real-world IoT prototype and further assessed through a sensitivity analysis and a computational study.