ContinuumConductor : Decentralized Process Mining on the Edge-Cloud Continuum

TL;DR

The paper addresses the challenge of applying process mining to distributed IIoT data with strict latency, privacy, and bandwidth constraints. It proposes ContinuumConductor, a 16-question decision framework that guides where each step of the process mining pipeline should run across the edge-cloud continuum, and demonstrates its use in an inland-port use-case (InteGreatDrones). Key contributions include a structured pipeline for IIoT process mining, an architecture-aware discussion of edge vs. cloud placement, and a practical framework to balance privacy, real-time responsiveness, and resource efficiency. The work lays a foundation for computing-aware, decentralized process mining in cyber-physical systems and invites further refinement and validation across scenarios.

Abstract

Process mining traditionally assumes centralized event data collection and analysis. However, modern Industrial Internet of Things systems increasingly operate over distributed, resource-constrained edge-cloud infrastructures. This paper proposes a structured approach for decentralizing process mining by enabling event data to be mined directly within the IoT systems edge-cloud continuum. We introduce ContinuumConductor a layered decision framework that guides when to perform process mining tasks such as preprocessing, correlation, and discovery centrally or decentrally. Thus, enabling privacy, responsive and resource-efficient process mining. For each step in the process mining pipeline, we analyze the trade-offs of decentralization versus centralization across these layers and propose decision criteria. We demonstrate ContinuumConductor at a real-world use-case of process optimazition in inland ports. Our contributions lay the foundation for computing-aware process mining in cyber-physical and IIoT systems.

Figures (2)

• Figure 1: Process Mining Pipeline. Raw sensor data is preprocessed to low-level events. These are aggregated to high-level events and correlated with case/object ids. Further, process models are discovered and insights are extracted from them.
• Figure 2: Process Mining Pipeline placed on the edge-cloud continuum. The preprocessing is performed distributedly on the edge nodes. On fog nodes, the data gets abstracted and correlated to cases. This results are then centrallized to perform discovery and insight extraction.