Method for implementing autonomous plant by encoding manufacturing domain knowledge in network of extensible and configurable ontologies

Abstract

It is therefore an object of the present invention to define a method of managing a single abstraction layer data model of different applications using semantic techniques and to automate the process of generating interoperability rules among different manufacturing sites. According to the invention, this object is achieved by a method for enabling an MOM system for an autonomous plant by encoding manufacturing domain knowledge in an equal-weighted co-control network of extensible and configurable ontologies, said method comprising the following steps: a) defining an ontology of the MOM domain by semantic conceptualization of manufacturing domain knowledge; b) representing the ontology by at least one mapping knowledge domain; c) implementing the ontology in an owl format; d) providing a web-based framework to provide a set of intuitive user-friendly user interfaces for navigating the ontology and associating concepts included in the ontology with logical data structures from existing applications included in the MOM system; e) mapping the logical data structures by semantic conceptualization semantic matching, f) generating interoperability codes between the data structures of the different applications, thereby mapping the data structures of the different applications within the same MOM system and / or over a plurality of distributed MOM systems of the autonomous plant.

Description

[0001] This invention relates to a method for realizing an autonomous factory by encoding manufacturing domain knowledge in a network of scalable and configurable ontology.

[0002] In the fields of process automation and process monitoring, standard automation systems for controlling the widest range of conceivable production resources, machines, and plants (MOM objects) represent the latest technology. This technology specifically encompasses the extensive product range offered by Siemens under its OpCenter product family in the field of Manufacturing Operations Management (MOM). A rich product line for addressing the technical tasks in question—such as counting, measuring, positioning, motion control, closed-loop control, and cam control—enhances the performance capabilities of appropriate process controllers. Various configurations support the implementation of flexible machine concepts.

[0003] In this context, a wide range of IT solutions exist to connect the physical hardware close to the technology and / or logistics processes to the application layer of the driving facility. Therefore, Manufacturing Execution Systems (MES) have been developed to meet all the requirements of Service-Oriented Architecture (SOA) for seamless integration into Fully Integrated Automation (TIA). The plug-and-play architecture, where individual functions can be configured and easily combined with each other, forms the basis of this success, thereby simplifying the complex structures controlling manufacturing plants, etc.

[0004] Therefore, manufacturing companies face various challenges, with the most prominent being increased variability across multiple dimensions. These dimensions include product characteristics, production volume, production processes, manufacturing technologies, customer behavior and supplier strategies, and even the transformation process of the entire industrial sector. This trend is accelerating, directly impacting the value chain and its associated physical supply chains, as well as factory design and management.

[0005] The prevailing view is that to meet these challenges and continue to remain competitive, companies need to embrace the digitization of all processes—not just production lines, but the entire value chain: from customer orders to product delivery, enabling seamless information flow. The impact of digitization on industrial processes is so significant that it is considered not merely an improvement, but a revolutionary step in industrial history. A specific term has been introduced to characterize it: Industry 4.0. The vision of Industry 4.0 is to transform traditional factories into smart factories, where processes are managed by software applications that seamlessly link business, production, logistics, and R&D within a holistic system. Siemens is responding to this trend by envisioning the Autonomous Factory concept, which focuses on highly flexible, resilient, sustainable, and self-organizing production for the future.

[0006] Of course, data is the foundation of this revolution. Data must be processed, exchanged, imported, displayed, and aggregated for manufacturing to run successfully.

[0007] A major challenge in data management in complex scenarios is that data must be processed in the same way by all participants, but it is often encoded in different formats. These data format differences are a significant obstacle to seamless application integration and increase the deployment costs of complex systems.

[0008] Furthermore, for software vendors, coordinating the data models of their applications is difficult, and sometimes it is also difficult to coordinate the data models between different versions of the same software. This is a significant problem that hinders innovation and increases development and maintenance costs.

[0009] Semantic technologies can offer a solution for mitigating problems and reducing costs. They associate semantic meaning with data structures in a way that is understandable to both humans and machines. For humans (whether users, developers, system integrators, or data analysts), this feature helps them understand information and improve productivity; for machines, it enables consistent processing of data with the same semantic meaning but in different formats.

[0010] Semantic modeling can facilitate data use and distribution because it allows for a single, semantically-based approach to managing data in different formats. A traditional method for managing data encoded in different formats is through a message broker, also known as enterprise application integration. This additional application transforms the data structure from one format to another. The application typically connects to the broker via a connector, which can be application-specific (i.e., specifically designed and implemented for the application) or generic (e.g., using standard formats). A typical example in industrial automation is Business-to-Manufacturing Markup Language (B2MML), a set of XML schemas that implements the data model in the ISA-95 standard.

[0011] Traditional methods have several constraints: connectors map application data models to internal models (canonical data models); at runtime, communication between applications requires two transformations: one for incoming data from the original data model to the canonical data model, and another for outgoing data from the canonical data model to the target model. Developing a connector requires complete knowledge of the application data model, and once developed, it is linked to a specific version of the model. This results in high creation costs, difficult maintenance, and poor performance.

[0012] Therefore, the object of this invention is to define a method for using semantic techniques to manage a single abstraction layer data model for different applications and, where necessary, to automate the process of generating interoperability rules.

[0013] According to the present invention, this objective is achieved by a method for enabling a MOM system for autonomous factories by encoding manufacturing domain knowledge in a democratic network of scalable and configurable ontologies.

[0014] The method includes the following steps:

[0015] a) Define the ontology of the MOM domain by semantically conceptualizing manufacturing domain knowledge;

[0016] b) The ontology is represented by at least one knowledge graph;

[0017] c) Implement the ontology in OWL format;

[0018] d) Provide a web-based framework to offer a set of intuitive, user-friendly user interfaces for navigating the ontology and associating the concepts included in the ontology with logical data structures from existing applications included in the MOM system;

[0019] e) Mapping logical data structures through semantic matching based on semantic conceptualization;

[0020] f) Generate interoperability code between data structures of different applications, thereby mapping data structures of different applications within the same MOM system and / or across multiple distributed MOM systems in an autonomous factory.

[0021] Therefore, this invention introduces a method for managing data at the conceptual level using MOM / MES ontology and the semantics employed. This method simplifies the work of users at all levels because processing concepts is easier for humans than processing data structures, and because the ontology can be processed by both humans and machines, it enables the automation of various processes. This method achieves a solution dedicated to automating multi-application data management in the field of manufacturing software.

[0022] Preferred embodiments of the present invention are given in appended claims 2 to 4.

[0023] In a preferred embodiment of the present invention, the body may include one or more of the following elements:

[0024] i) Concept

[0025] ii) Terms used to identify concepts, preferably consisting of vocabulary and industry-specific terms.

[0026] iii) Semantic description of concepts

[0027] iv) Attributes

[0028] v) Verbs, especially those indicating actions that can be performed on a concept by a command.

[0029] vi) Relationships, especially relationships between concepts

[0030] vii) Rules, such as rules used for data transformation, data processing, and data validation.

[0031] To automatically handle data exchange in various applications, ontology can be implemented using the RDF format (RDF = Resource Description Frame), where a concept is represented as a set of so-called "triples," which are three entities that encode a statement about semantic data in the form of a subject-verb-object expression.

[0032] In addition, the following RDF fields can be filled in to express the concept:

[0033] i) Label: Used for conceptual terms

[0034] ii) Comments: Used for semantic description

[0035] iii) IsDefinedBy: Used for domain definition

[0036] iv) Class / Subclass: Used for conceptual hierarchy structure

[0037] v) Individual: Used for industry terminology association

[0038] vi) Type: Used to associate concepts with data structures

[0039] vii) ObjectProperty: Used for defining people and relationships.

[0040] The preferred embodiments are described in more detail below with reference to the accompanying drawings, which are depicted as follows:

[0041] Figure 1 This schematically illustrates the organization of semantic catalog information for the purpose of defining the data ontology in a MOM system;

[0042] Figure 2 An example of a concept related to the MOM domain is illustrated schematically;

[0043] Figure 3 An example of a knowledge graph for device concepts in the MOM ontology is illustrated schematically;

[0044] Figure 4 The RDF description of the "EquipmentClass" concept is illustrated schematically.

[0045] Figure 5 This schematically illustrates the abstract relationship from the MOM ontology to the physical data model in the MOM system;

[0046] Figure 6 The illustration schematically shows several operations supported in the UI that enable users to access, query, and navigate ontology and knowledge graph;

[0047] Figure 7 This schematically illustrates an example of C# code automatically generated using Microsoft's Roslyn.NET compiler, starting from a concept diagram; and

[0048] Figure 8 This illustration shows an example of the application of ontology concepts when sharing data between two factories, A and B.

[0049] This invention is part of a MES / MOM solution, such as the one launched by Siemens AG under the trademark name Opcenter EX FN, and aims to deliver greater value to multi-plant solutions by improving data operability across different applications and systems / tools / resources in a manufacturing entity.

[0050] Specifically, this invention defines a method for using semantic techniques to manage a single abstraction layer data model for different applications and, where necessary, automating the process of generating interoperability rules. Writing interoperability code between applications has always been one of the most costly and error-prone activities in developing distributed systems, such as modern MOM systems. Now, this process can be automated starting with a semantic description of the information that apps must exchange. As a first step, it is necessary to define the correspondence between the concepts available in the semantic model and the corresponding data models of the systems that must interoperate. Following this configuration, the system can automatically generate the necessary code to enable apps to exchange information.

[0051] The advantages of this method are that it reduces manual design time, accelerates development, and reduces errors. To enable user-friendly use and management of the ontology, the invention also includes a web application with several user interfaces that allow users to easily access, query, and navigate the ontology and knowledge graph.

[0052] In the domain of Manufacturing Execution Systems (Manufacturing Operations Management), an ontology encompasses the representation, formal naming, and definition of categories, attributes, and relationships between concepts, data, and entities that concretize one, several, or all domains of discussion. This ontology concept is rigorously specified by the ANSI / ISA-95 standard. Therefore, an ontology demonstrates the attributes of a subject domain (e.g., an autonomous factory) and how they relate to each other by defining a set of concepts and categories that represent a subject (plant).

[0053] Contemporary ontologies from different manufacturing sites, regardless of the languages ​​used to express them, share many structural similarities. Most ontologies describe individuals (instances), classes (concepts), attributes, and relationships. Each of these components is briefly discussed below.

[0054] Common components of an ontology include:

[0055] a) Individual

[0056] Instances or objects (basic or "underlying" objects)

[0057] b) Class

[0058] Set, collection, concept, class in programming, object type or kind of thing

[0059] c) Attributes

[0060] Objects (and classes) can have aspects, attributes, characteristics, properties, or parameters.

[0061] d) Relationship

[0062] The ways in which classes and individuals can be related to each other

[0063] e) Function Items

[0064] A complex structure formed by certain relationships, which can be used in place of individual items in a statement.

[0065] f) Restrictions

[0066] For an assertion to be accepted as input, it must be a formal declarative description that is true.

[0067] g) Rules

[0068] A statement in the form of an if-then (premise-result) sentence describes a logical inference that can be derived from a particular form of assertion.

[0069] h) Axioms

[0070] Assertions (including rules) in logical form together constitute the overall theory describing an ontology within its domain of application. This definition differs from the definition of "axiom" in generative grammar and formal logic. In these disciplines, axioms only include statements asserted as a priori knowledge. As used herein, "axiom" also includes theories derived from axiomatic statements.

[0071] i) Event

[0072] Changes in attributes or relationships

[0073] Ontologies are typically coded using an ontology language (i.e., OWL files). An ontology editor is an application designed to assist in the creation or manipulation of ontology. Ontology editors typically use one or more ontology languages. Aspects of an ontology editor include: visual navigation possibilities within the domain knowledge model, inference engines, and information extraction; support for modules; importing and exporting external knowledge representation languages ​​for ontology matching; and support for meta-ontologies (such as OWL-S, Dublin Core, etc.).

[0074] This invention defines a method for managing data models of different apps in a MOM domain using ontology and knowledge graphs. It provides a complete, user-friendly framework to support ontology navigation, the association of semantic concepts with application data structures, the creation of interoperability rules between different applications, and the automatic generation of API code for interoperability.

[0075] The main concept of this invention is:

[0076] - Define the ontology of the MOM domain through semantic conceptualization.

[0077] - Representing the ontology using a knowledge graph

[0078] - Implement the ontology in owl format

[0079] - Implement a web-based framework to provide a user-friendly UI for navigating the ontology and associating concepts with logical data structures from existing applications.

[0080] - Mapping data structures through semantics

[0081] - Automatically generate interoperability code between data structures of different applications.

[0082] Definition and Representation of MOM Knowledge Base Based on Semantic Graph

[0083] In information technology, the representation of a knowledge base for an application domain is the domain's ontology. A dictionary defines concepts, semantic descriptions of images, and relationships between concepts (taxonomy). It can be visually represented as a knowledge graph, where images are nodes and connections are links between nodes.

[0084] As mentioned above, an ontology can consist of the following elements:

[0085] -concept

[0086] - Terms used to identify concepts; these consist of vocabulary and industry-specific terminology.

[0087] - Semantic description of the concept

[0088] -property

[0089] - Verbs (actions that a command can perform on a concept)

[0090] - Relationships (between concepts)

[0091] - Rules (transformation, processing, and validation)

[0092] In the MOM application domain, these concepts are manufacturing-related. Here, the ISA-95 standard is used as a reference for the vocabulary of images. ISA-95 provides a consistent terminology to support communication in manufacturing operations and provides an operational model that interprets the semantics of the language.

[0093] The MOM application domain is further divided into subdomains based on operation type (i.e., manufacturing execution, quality assessment, logistics, etc.) and application type (e.g., automotive, aerospace, CFG, etc.) to better focus on operational conditions.

[0094] under Figure 1 This paper summarizes the organization of information in the MOM ontology.

[0095] The global MOM domain can be grouped within the self-consistent operation domain, such as execution, quality assurance, intraplant logistics, and warehousing. Each subdomain can be further subdivided based on manufacturing type (e.g., discrete, process, pharmaceutical). Figure 1 As shown, the domain itself is composed of concepts and relationships. Images have semantic associations, terms (names), and some attributes. Depending on the application implementation, a single concept can be associated with some industry terms. Finally, the domain has a software implementation that is an application of the Opcenter family (Execution Core, ExecutionDiscrete, Modular MOM, etc.).

[0096] Here, the ontology is implemented using RDF (Resource Description Frame). In RDF, a concept is represented as a set of so-called "triples," which are three entities that encode a statement about semantic data in the form of a subject-verb-object expression. Specifically, the following RDF fields are populated:

[0097] Label: Used for conceptual terms

[0098] Comments: Used for semantic descriptions

[0099] IsDefinedBy: Used for domain definition

[0100] Class / Subclass: Used for conceptual hierarchy structures

[0101] Individual: Used for industry terminology association

[0102] Type: Used to associate concepts with data structures

[0103] Object Property: Used for defining people and relationships.

[0104] The complete ontology is provided as an OWL file (OWL = W3C Web Ontology Language). Semantic queries of the catalog can be performed using SPARQL syntax. Figure 2 Below are examples of implementing knowledge bases and knowledge graphs.

[0105] Figure 3 This diagram illustrates an example of a knowledge graph for device concepts within the MOM (Manager of Machines) entity. Light gray dots represent concepts, dark gray dots represent application industry terminology, the upper right represents industry terminology for Execution Discrete, and the upper left represents industry terminology for Opcenter Execution Core. Accordingly, Figure 1 This schematically represents the RDF description of the "Equipment class" concept.

[0106] Mapping ontology with logical application model

[0107] An ontology represents the formalization of a knowledge base at an abstract level. For it to be machine-readable, it must first be transformed into a logical data model in the form of data structures, and finally into a physical data model in the form of tables and records. An ontology is an abstract formalization of a knowledge base associated with an application domain, possessing temporal consistency and generality; the logical data model is a formalization of data structures associated with the application's software version and can vary according to performance; the physical model is a data organization associated with the software build and is build-specific.

[0108] The final step is performed by the compiler of the computer language used by the application. The first step can be performed directly or indirectly through software tools.

[0109] Direct transformation is performed to create a new application logic model from the ontology. Direct transformation creates new data structures for the analysis model from the ontology's classes and subclasses; several tools exist that can perform this step. Indirect transformation associates ontology concepts with the data structures of an existing application's implemented logic model.

[0110] This is the focus of the invention because such a tool does not exist in the market. This tool enables the management of the logical models of one or more existing applications through semantics and knowledge graphs, thereby adding the advantages of semantic management without requiring changes to the analysis model. Therefore, Figure 5 It represents the abstract relationship from the MOM ontology to the physical data model existing in the corresponding MOM system.

[0111] Part of this invention is a web application that loads the ontology and logical data model of one or more applications, and performs automatic association between concepts in the ontology and data structures in the analysis model. The web application also provides at least one UI for navigating the ontology, data model, and then manipulating that information. The web application will be described in detail below.

[0112] The mapping between the application data model and the conceptual model in the ontology is performed by a likelihood algorithm, which provides the association between entities in the application data model and the structure of the conceptual model (concept graph). In this example, the likelihood algorithm is based on string distance and implements three types of distance (Hamming distance, Lewinstein distance, and Demero-Lewinstein distance). Lewinstein distance is a string metric used to measure the difference between two character sequences. The Lewinstein distance between two words is the minimum number of single-character edits (i.e., insertions, deletions, or replacements) required to change one word into another. It is named after Vladimir Lewinstein. The likelihood algorithm enables the broadening of likelihood associations using industry terminology associated with the same concept.

[0113] Likelihood algorithms also provide confidence levels for automatic mapping. Relationships can be of types 1:1, 1:N, N:1, and N:M. Typically, a concept graph exists for each application requiring interoperability. Mappings are presented to the user for validation via a visual interface (UI). Users can confirm the automatic mapping or make manual adjustments. Users can add industry terminology to the data model to facilitate future associations.

[0114] User-friendly web framework for knowledge base representation

[0115] To enable the use and management of ontology, a framework has been implemented that creates web applications with several user interfaces (UIs) that allow users to access, query, and navigate ontology and knowledge graph. Figure 6Several operations supported in the UI are illustrated as examples.

[0116] The main goal of the user interface is to provide a natural and intuitive way to explore the ontology, while shielding the user from technical aspects such as performing SPARQL queries.

[0117] Users begin by selecting an ontology model. They then search for concepts of interest using standard terminology, such as "job orders."

[0118] Then, the user is presented with a list of concepts that match his / her search. These images are listed in order from most relevant to least relevant. Each concept is an OWL class.

[0119] After selecting a concept from the list, the user is presented with a list of entities grouped by "domain", where each entity is an OWL named instance and the domain is the RDF isDefinedBy string.

[0120] Users select zero or one entity from each domain and are presented with a visualization of the specified entity.

[0121] The visualization is presented as a force-directed graph of nodes (classes and entities) connected by edges (relationships between entities). Initially, selected entities are shown as those connected to their immediate next-level neighbors. This initial view can be configured by adjusting the "Depth" parameter. Nodes can be expanded or collapsed to expose / hide most of the ontology.

[0122] After clicking on a project, a pop-up window will display detailed information about each class, entity, and connection (such as description, type, and URI).

[0123] Automatic generation of interoperability code and automatic API creation

[0124] The relationship between the ontology concepts defined in the preceding sections and the logical data structures of applications enables the automatic generation of interoperability rules to exchange consistent data across different applications. In fact, if two structures in two different application logical models refer to the same concept, it is easy to find mapping rules for passing attributes. When the relationship between concept and data structure is 1:1, interoperability mapping can be performed automatically. In cases involving N:1, 1:N, or N:N relationships, manual verification is required to resolve conflicts.

[0125] The mapping is performed in the web application tools, and the API generation is provided in the form of C# code and DLL binaries. Automatic code generation is done using the Microsoft Roslyn.NET compiler.

[0126] Figure 2An example is shown of C# code automatically generated using the Roslyn.NET compiler, starting from a concept diagram; along with the code, an associated binary DLL is also automatically generated.

[0127] Application Examples

[0128] This section provides practical examples of how the invention can generate value for typical industrial applications.

[0129] Cross-industry information management

[0130] Manufacturing execution exhibits several differences and characteristics depending on the industry segment in which the application is performed (i.e., discrete manufacturing, process manufacturing, food and beverage, life sciences, complex manufacturing, etc.). Other differences may arise from national regulations, customer-specific conditions, etc. Figure 1 The organization of information presented here is an efficient and flexible information authorization mechanism. The fundamental information of the application domain can be separated through different re-representations. At the top is a general ontology valid for each industry segment and condition. Industry-specific information can be represented in element-specific ontology, which can be associated as an extension of the public ontology to form a complete industry-domain ontology. This representation can be further iterated, for example, through country-specific and customer-specific ontology, to obtain a complete formalization of the customer knowledge base, and reuse much of the general knowledge base.

[0131] A practical example could be the definition of operations. Procedural operations are defined at the general ontology level, operation sequences are determined at the sub-domain ontology level, and operation types can be defined at the client-specific ontology level. In this way, most of the information can be reused, requiring only client-specific extensions to be implemented for each new project.

[0132] Interoperability in distributed layered systems

[0133] Customers can operate in a distributed, hierarchical environment where orders are managed at a central level, while execution takes place in several geographically distributed facilities / factories. Each facility may have a slightly different manufacturing management system; for example, different releases of a given product, or entirely different effects. In this scenario, the flow of information from the central system to the final design can be complex and require specialized modifications. These difficulties are addressed through ontology-based information management and the automatic generation of interoperability rules.

[0134] like Figure 6 As shown, the ontology ensures that data points to the same semantic meaning across all organizational data in different systems, thereby enabling... Figure 7 The automatic generation of the interoperability rules shown in the diagram is made possible, and thus enables information to flow seamlessly throughout the hierarchical system.

[0135] Specific example of generating interoperability rules for two factories A and B

[0136] Figure 8 A specific example of interoperability rule generation for two factories, A and B, is shown. Factory A is producing canned tomato soup and uses an MES from a first MES software provider (e.g., Siemens and its Opcenter software suite). Factory B is producing convenience foods for pasta dishes with different sauces (e.g., tomato sauce or vegetable sauce). Factory B is controlled by an MES from a second MES software provider. Now assume that Factory B should exchange information with Factory A, for example, at the management level of procuring various raw materials used for their respective products. It is clear from this specific example that the two factories require some of the same raw materials, such as tomatoes, herbs, spices, water, etc. Therefore, the two factories A and B wish to exchange information about the consumption of these raw materials in their respective production. Consequently, the resources used in these production processes also have a similar range, such as mixers, heaters, metering devices, packaging, etc. Therefore, it may also be expected that the ERP system receives some information about the use and energy consumption of these resources to identify, for example, KPIs for specific production batches.

[0137] To solve this task, Figure 8 This describes the activity flow according to the present invention. In step 1, typical MOM / MES ontology and relationships for this segment of the food and beverage industry are provided to factory A almost out-of-the-box. In step 2, the concepts of the MOM / MES ontology are then linked to entities managed by a second MES system in factory B. In step 3, a code generation engine automatically generates interoperability code, enabling the two MES systems to exchange data within their respective entities without requiring users to write / develop their own code. In step 4, during runtime, the interoperability code is used to ultimately enable interaction between the two MES systems, thereby facilitating data exchange, for example, regarding raw material consumption, remaining raw material inventory, energy consumption, maintenance activities, etc.

[0138] Therefore, this invention introduces a method for managing data at the conceptual level using MOM / MES ontology and the semantics employed. This method simplifies the work of users at all levels because humans find it easier to process concepts than data structures, and because the ontology can be processed by both humans and machines, it enables the automation of various processes. This method achieves a solution dedicated to automating multi-application data management in the field of manufacturing software.

[0139] Compared to the mainstream solutions mentioned in the introduction, the main advantages are as follows:

[0140] This approach enables conceptual management of data across multiple application environments. While data may have specific implementations in different applications, it can be managed in a unified manner. It provides a machine-readable abstract model that can be automatically transformed into an application data model. Furthermore, the approach offers user-friendly tools for navigating the ontology. It also provides automated code generation to achieve interoperability between applications within the same domain but with different data models. Therefore, it reduces development time and effort for both development teams and clients / users.

[0141] Siemens is a leading provider of Manufacturing Operations Management (MOM) systems, offering diverse suites covering all segments of the MOM market. Because its product portfolio was built through acquisitions of different companies and the integration of heterogeneous products, Siemens has experienced the traditional difficulties of integrating different data models. The use of ontology and semantics can facilitate data management for applications, thereby providing better solutions by reducing integration costs. Furthermore, semantic data management can facilitate a balance between new products and mature applications within the product portfolio. Finally, semantic data management can facilitate integration with third-party products, enhancing product value while reducing costs and creating more value for its customers.

Claims

1. A method for enabling MOM systems for autonomous factories by encoding manufacturing domain knowledge in a weighted, co-controlled network of scalable and configurable ontology. The method includes the following steps: a) Define the ontology of the MOM domain by semantically conceptualizing the knowledge of the manufacturing domain; b) The ontology is represented by at least one knowledge graph; c) Implement the ontology in owl format; d) Provide a web-based framework to offer a set of intuitive, user-friendly user interfaces for navigating the ontology and associating the concepts included in the ontology with logical data structures from existing applications included in the MOM system; e) Mapping the logical data structure through semantic matching based on the semantic conceptualization; f) Generate interoperability code between data structures of different applications, thereby mapping the data structures of the different applications within the same MOM system and / or across multiple distributed MOM systems in the autonomous factory.

2. The method according to claim 1, wherein, An ontology includes one or more of the following elements: i) Concept ii) Terms used to identify the concept, preferably consisting of vocabulary and industry-specific terms. iii) Semantic description of the concept iv) Attributes v) Verbs, especially those expressing the actions that a command can perform on a concept. vi) Relationships, especially relationships between concepts vii) Rules, such as rules used for data transformation, data processing, and data validation.

3. The method according to claim 1 or 2, wherein, The ontology is implemented using the RDF format (RDF = Resource Description Frame), where the concept is represented as a set of so-called "triples," which are three entities that encode a statement about semantic data in the form of a subject-verb-object expression.

4. The method according to any one of the preceding claims, wherein, Populate the following RDF fields to express the concept: i) Labels: used for conceptual terms ii) Notes: used for semantic description iii) IsDefinedBy: Used for domain definition iv) Classes / Subclasses: Used for conceptual hierarchy structures v) Individual: Used for industry terminology association vi) Type: Used for the concept-to-data-structure association vii) Object properties: used for defining people and relationships.

Images