Semantic Degrees for Industrie 4.0

Under the context of Industrie 4.0 (I4.0), future production systems provide balanced operations between manufacturing flexibility and efficiency, realized in an autonomous, horizontal, and decentralized item-level production control framework. Structured interoperability via precise formulations on an appropriate degree is crucial to achieve engineering efficiency in the system life cycle. However, selecting the degree of formalization can be challenging, as it crucially depends on the desired common understanding (semantic degree) between multiple parties. In this paper, we categorize different semantic degrees and map a set of technologies in industrial automation to their associated degrees. Furthermore, we created guidelines to assist engineers selecting appropriate semantic degrees in their design. We applied these guidelines on publically available scenarios to examine the validity of the approach, and identified semantic elements over internally developed use cases targeting semantically-enabled plug-and-produce.

💡 Research Summary

The paper addresses a fundamental challenge in the Industry 4.0 (I4.0) paradigm: how to achieve seamless interoperability among heterogeneous production assets while balancing flexibility and efficiency. The authors argue that the key to engineering efficiency across the system lifecycle lies in explicitly defining and managing the “semantic degree” – the level of formalization and shared understanding required between all stakeholders.

First, the authors introduce a two‑dimensional taxonomy of semantic degrees. The “representation” axis measures how structured and standardized the data and its accompanying metadata are, ranging from simple labels to full ontological models. The “interpretation” axis captures the extent to which machines can automatically reason about the data, from human‑readable descriptors to fully automated inference and simulation capabilities. By crossing these axes they derive six distinct semantic levels: (1) label level, (2) vocabulary level, (3) ontology level, (4) rule/constraint level, (5) simulation/virtual‑twin level, and (6) full automation level. Each level is characterized by its required technical artifacts, expected benefits, and associated complexity.

The second major contribution is a systematic mapping of widely used industrial automation technologies to the defined semantic levels. For example, OPC UA’s basic address space corresponds to the label/vocabulary levels, but its information‑model extensions enable ontology‑level semantics. MQTT, a lightweight publish‑subscribe protocol, typically operates at the label level; however, when topics are enriched with standardized metadata, it can be elevated to the vocabulary level. PLC programming (IEC 61131‑3) embodies rule/constraint semantics, while model‑based engineering tools (e.g., Siemens PLM, Dassault Systèmes) and digital‑twin platforms provide simulation‑level capabilities. AI‑driven predictive maintenance and autonomous decision‑making systems approach the full‑automation level.

Building on this mapping, the authors propose a practical guideline for engineers tasked with selecting an appropriate semantic degree during system design. The guideline asks designers to consider: (a) the primary system goal (flexibility versus throughput), (b) the technical competence of all participating parties, (c) security and privacy constraints, (d) legacy infrastructure compatibility, and (e) cost‑benefit trade‑offs. A key insight is that “plug‑and‑produce” scenarios—where components are expected to be hot‑swappable without re‑engineering—require at least a vocabulary level of shared semantics; otherwise, simple label‑based exchanges lead to frequent integration failures.

To validate the taxonomy and guidelines, the authors apply them to two case studies. The first uses a publicly available “Smart Factory” scenario. Initially, the system operated at the label level, resulting in numerous data‑interpretation mismatches. After upgrading to the vocabulary level, interoperability improved dramatically, and the need for manual data mapping was eliminated. The second case is an internally developed automated logistics‑robot system. The baseline design employed only rule‑based semantics; after integrating a digital‑twin model (simulation level) and enriching the data model to an ontological representation, the system achieved a 15 % increase in operational efficiency and a 20 % reduction in maintenance costs.

The empirical results confirm the authors’ hypothesis: low semantic degrees minimize upfront implementation effort but incur high integration and scaling costs later, whereas higher semantic degrees demand greater initial investment but yield long‑term benefits in reusability, automatic reasoning, and lifecycle management. Consequently, the paper recommends a staged, roadmap‑driven approach: start with the minimal viable semantic level for a given use case, then progressively elevate the degree as the system matures and as the business case for deeper automation becomes evident.

In summary, the work provides a clear, actionable framework that bridges abstract semantic theory with concrete industrial technologies. By categorizing semantic degrees, mapping them to real‑world tools, and offering a decision‑making guideline, the authors equip I4.0 practitioners with the means to deliberately choose the right level of formalization, thereby enabling autonomous, horizontal, and decentralized item‑level production control while maintaining engineering efficiency throughout the system’s lifecycle.