A planning and lifecycle cost model for industrial Internet of Things systems

Industrial Internet of Things (IIoT) is a pillar of digital transition and Industry 4.0 and 5.0. However, effective IIoT planning requires matching architectures to business needs alongside reliable cost estimation, but decision-making tools in the literature are scarce. Existing cost models are narrow in scope, frequently focus on specific applications, neglect relevant costs and uncertainties, while few guidelines exist to select proper IIoT architectures. To contribute to fill this gap, this paper develops a decision support framework for strategic planning and lifecycle cost estimation across five architectures: On-Premise, Cloud-Based, Service-Oriented, Modular, and Hybrid. The methodology allows qualitative architecture preselection, novel comprehensive parametric lifecycle cost modeling with Monte Carlo uncertainty propagation, including cyber-risk penalties, and a Fuzzy TOPSIS multi-criteria decision-making for final selection in an unified framework which was previously missing. The methodological novelty lies in the structural integration of these elements, where cost-based stochastic risk propagation, is directly incorporated into the multi-criteria decision environment, mathematically binding quantitative risk outputs to qualitative expert judgments. An illustrative example demonstrates the framework's applicability. Findings identify major cost items impacting each architecture, show that propagating inherent uncertainty can overturn deterministic results. Under the assumptions of the illustrative scenario, Cloud-Based configuration minimizes initial investment but exhibits highest exposure to cloud-price volatility and cyber-related tail risk. On-Premise configuration shows lowest extreme-cost exposure, despite higher upfront capital intensity. Hybrid configuration yields intermediate average costs but remains exposed to local infrastructure uncertainty and residual external-dependency risk.