Catch up on the latest AI articles

ExtruOnt: An Ontology For Describing Types Of Manufacturing Machinery For Industry 4.0 Systems

ExtruOnt: An Ontology For Describing Types Of Manufacturing Machinery For Industry 4.0 Systems


3 main points

✔️ A semantic description of manufacturing machinery is needed for Industry 4.0, but is not currently in place.
✔️ This paper addressed the development of an ontology for one example.
It contains classes and properties to represent the description of the equipmentcomponents, spatial connections, features, 3D representation of the components, and finally the sensors used to obtain indicators about the performance of this kind of machine.

ExtruOnt: An ontology for describing a type of manufacturing machine for Industry 4.0 systems
written by Víctor Julio Ramírez-DuránIdoia BergesArantza Illarramendi
[Submitted on 22 Jan 2024]
Comments: This is the accepted manuscript. The definitive, peer reviewed and edited version of this article is published in Semantic Web 11(6): 887-909 (2020)
  Artificial Intelligence (cs.AI)


The images used in this article are from the paper, the introductory slides, or were created based on them.


A semantically rich description of manufacturing machinery, provided in machine-interpretable code, can be used effectively in an Industry 4.0 scenario. However, such a description is clearly lacking. This paper presents a development effort to build an ontology, called ExtruOnt, to describe a type of manufacturing machinery, more precisely, a type that performs extrusion processes (extruder). Although the scope of this ontology is limited to a concrete domain, it can be used as a model for the development of other ontologies to describe manufacturing machines in an Industry 4.0 scenario. the terms in the ExtruOnt ontology are information about different types of extruder-related and they are reflected in the individual modules that make up the ontology. Thus, it contains classes and properties to represent descriptions about the components of the extruder, spatial connections, features, 3D representations of the components, and finally sensors used to obtain indicators about the performance of this type of machine. The process of developing the ontology was carried out in close cooperation with experts in the respective fields.


Various initiatives and strategies are emerging in the manufacturing industry, known as the Fourth Industrial Revolution (Industry 4.0). These initiatives aim to collect data on product history, condition, quality, and characteristics, and apply manufacturing intelligence that leverages this data. This is creating important business opportunities for manufacturers.

Proper design and implementation of such initiatives will require innovative efforts in mechatronics, manufacturing strategies, knowledge workers, and the use of modeling, simulation, and forecasting methods and tools.

In particular, it has been noted that from a modeling perspective, there is a lack of adequate descriptions of manufacturing machinery that are accessible, interoperable, and reusable. The authors have therefore developed ExtruOnt, an ontology that provides a detailed description of a real manufacturing machine type called extruder.

The ExtruOnt ontology contains terms for the main components of an extruder, their spatial connections, features, 3D representations, and sensors that capture the performance of this type of machine.ExtruOnt is implemented using the OWL 2 Web ontology language and is built in Protégé development environment.

ExtruOnt is consistent with existing ontologies such as the DUL ontology, MASON ontology, and SAREF4INMA. It also reuses terminology from ontologies such as GeoSPARQL, OM, and 3DMO.

Related Research

Several ontologies related to the manufacturing sector can be found in the literature. These ontologies are defined for different purposes and describe different types of information related to the field.

The PSL ontology contains the basic concepts for representing manufacturing processes. It defines concepts such as activity, activity occurrence, time point, and object as the basic elements of a manufacturing process.

The MASON ontology is a high-level ontology for representing the core concepts (products, processes, and resources) of a manufacturing domain. It defines the basic classes of products, manufacturing operations, and manufacturing resources.

The SIMPM ontology is a high-level ontology that models the basic constraints of manufacturing process planning. It includes concepts related to manufacturing activities, resources, time, and aggregation.

The MaRCO ontology defines the functions of manufacturing resources. Classes are defined to represent single functions (e.g., fixing, cutting) and compound functions (e.g., pick and place, move and release).

The MSDL ontology enables the description of manufacturing services. It includes concepts such as manufacturing services, providers, manufacturing capabilities, manufacturing resources, and manufacturing processes.

The P-PSO ontology considers three aspects of the manufacturing domain: physical, technical, and control. Concepts such as components, operations, and controllers are defined.

The OntoSTEP ontology primarily enables the description of geometric information about a product. It can describe the shape, dimensions, location, etc. of a product.

The MCCO Ontology focuses on the interoperability of the design and production areas of the product lifecycle. It includes concepts such as manufacturing processes, manufacturing facilities, manufacturing resources, and features.

The SAREF4INMA ontology is intended to promote interoperability with industrial standards. Concepts such as manufacturing equipment, plant, product and material categories are defined.

While some of these ontologies contain general industrial machinery concepts, further specialization and characterization is needed to describe and characterize specific industrial machinery types in detail; the ExtruOnt ontology was built for this purpose.

Development of the ExtruOnt ontology

The ExtruOnt ontology was developed using the NeOn methodology, which was determined to be the best fit for ExtruOnt's requirements because it considers different ontology construction scenarios and provides detailed guidelines for ontology construction activities.

The ExtruOnt development process follows the NeOn methodology's 6-phase + merging phase waterfall ontology network life cycle model (Figure 1). An overview of each phase is as follows

Figure 1. 6-phase + merge phase Waterfall Ontology Network Life cycle model and scenario, shown with activities and ExtruOnt module.

Launch phase

An Ontology Requirements Specification (ORSD) was developed to define the purpose, scope, and competency questions for the ExtruOnt The ORSD describes the purpose, scope, target users, intended use, and functional requirements (competency questions) of the ExtruOnt. The competency questions were categorized into five groups: questions about the components of the extruder, spatial connections between components, characteristics of the components, 3D representation of the components, and sensors that capture the performance of the components.

Reuse phase

Existing ontological and non-ontological resources were searched and used to build each module of ExtruOnt. We utilized literature and for the description of components, GeoSPARQL for the representation of spatial relationships, OM ontology for the description of features, X3D/3DMO for 3D representation, and SOSA/SSN ontology for the description of sensors.

Merging phase

Alignment was done to align with higher ontologies such as DUL, MASON, SAREF4INMA, etc. ExtruOnt reuses concepts from these higher ontologies to ensure interoperability.

Redesign Phase

Conceptual models were extracted from non-ontological resources and converted to an ontology. Knowledge gained from the literature about extruder components, features, and sensors was defined as ontology concepts.


Modularity facilitates ontology development, reuse, and maintenance. It also conforms to the five dimensional approach derived from the ORSD analysis. Therefore, ExtruOnt consists of five modules. This modular approach makes ExtruOnt a flexible and extensible ontology. Each module can be developed, reused, and maintained independently and can be used to develop ontologies for new manufacturing machine types.

implementation phase

Written in OWL 2 DL and implemented in Protégé, ExtruOnt consists of five modules.


The Maintenance Phase is currently underway. If an error is detected, it will be corrected back in the design phase according to the Waterfall Ontology Network Lifecycle Model. This maintenance process ensures that ExtruOnt is always up-to-date and flexible enough to adapt to changes in the manufacturing industry. Ongoing collaboration with domain experts also ensures that ExtruOnt is enhanced and improved.

Thus, the ExtruOnt development process followed the guidelines of the NeOn methodology, and each module of ExtruOnt was designed and implemented while making maximum use of existing ontology and domain knowledge. This development process has resulted in ExtruOnt being a design-quality ontology with adequate coverage of the extruder domain.

Modules of the ExtruOnt ontology

The ExtruOnt ontology consists of five modules to provide a detailed description of an extruder. These modules are designed to represent different aspects of an extruder.

Figure 2. ExtruOnt ontology diagram showing reuse of terms from other domain ontologies


This module represents the major components of an extruder. Specifically, it includes concepts such as drive system, feed system, screw/barrel/heating system, head/die assembly, and control system. The relationships between these components are defined using the PartOf ontology design pattern. The module is also consistent with higher-level ontologies such as SAREF4INMA's ProductEquipment and MASON's Machine-tool.

Figure 3. Components of the extruder.


This module represents spatial relationships between extruder components; in addition to standard spatial relationships such as RCC8 relationships, custom spatial relationship properties are defined. The GeoSPARQL ontology is utilized to represent these spatial relationships. Consistency with DUL's PhysicalObject is also ensured.

Figure 7. relationship between RCC5 and RCC8.


This module represents the characteristics of the extruder's components. Specifically, it contains information on dimensions, operating conditions, and productivity. Parts of the OM ontology are reused in the representation of these features.

Figure 11. Example definition of characteristic motor voltage measurements    


This module represents the 3D model and location of the extruder components; the 3DMO ontology is reused to provide a detailed 3D representation of the extruder.

Figure 13. 3D view of extruder components.


This module represents the sensors that capture extruder performance and their observed data; the SOSA/SSN ontology is reused and features of the extruder domain are added.

Figure 15. excerpt from the sensors4ExtruOnt module showing some classes and properties related to sensors.

These modules are designed to adequately represent the characteristics of the extruder domain while maximizing the use of existing ontologies. This modular approach makes ExtruOnt a flexible and extensible ontology.


The evaluation was conducted from two perspectives: domain coverage and design quality. For domain coverage, we conducted an exhaustive analysis of the literature material to determine the extent of ExtruOnt's coverage. For design quality, we used methods such as ontology metrics, detection of common design pitfalls, and application of evaluation criteria defined during the development process. The evaluation was also performed by three different stakeholders (R&D director of an ExtruOnt machine manufacturer, an IBDS provider, and an ontology expert).


This paper introduced the ExtruOnt ontology, which provides a detailed description of manufacturing machines in an Industry 4.0 scenario.ExtruOnt provides a reference model for describing the physical representation of a specific manufacturing machine type called an extruder and the data collected from its sensors.ExtruOnt contains information about the components of an extruder, their spatial connections, features, 3D representations, and sensors that capture the performance of this type of machine. The ontology development process was carried out in close collaboration with domain experts The main contributions of ExtruOnt are reusability, representativeness of spatial connections, and application to ontology-based systems ExtruOnt can be used as a basis for ontology development for other manufacturing machine types as well as for the development of ontology-based visual query and recommendation systems.Future challenges include continued maintenance and the development of software artifacts with ExtruOnt at its core.

  • メルマガ登録(ver
  • ライター
  • エンジニア_大募集!!
友安 昌幸 (Masayuki Tomoyasu) avatar
JDLA G certificate 2020#2, E certificate2021#1 Japan Society of Data Scientists, DS Certificate Japan Society for Innovation Fusion, DX Certification Expert Amiko Consulting LLC, CEO

If you have any suggestions for improvement of the content of the article,
please contact the AI-SCHOLAR editorial team through the contact form.

Contact Us