Systems Engineering (SE) is a vast discipline that includes many sub-disciplines. The International Council on Systems Engineering (INCOSE) defines Systems Engineering as:
“<..> an interdisciplinary approach and means to enable the realization of successful systems. It focuses on defining customer needs and required functionality early in the development cycle, documenting requirements, then proceeding with design synthesis and system validation while considering the complete problem: 

Operations

Cost and Schedule

Performance

Training and Support

Test

Disposal

Manufacturing

 Systems Engineering integrates all the disciplines and specialty groups into a team effort forming a structured development process that proceeds from concept to production to operation. Systems Engineering considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.“
Systems Engineering is a diverse and extensive field. Systems Engineers typically work for manufacturing and aerospace companies. They get involved in creating some of the most complex products built by man. In many ways Systems Engineering shares the use cases that motivated the development of Semantic Web technologies. Two leading standards support Systems Engineering:

1. OMG SysML™ the Systems Modeling Language, a standard graphical notation for systems modeling from the Object Management Group (OMG),  and
2. AP233 – a broad data interchange standard ISO 10303-233 Industrial automation systems and integration — Product data representation and exchange — Part 233: Application protocol: Systems engineering.

Both standardization activities were initiated by the INCOSE Model Driven Systems Design (MDSD) working group with the intent to keep the OMG and ISO standards aligned. With these standards in place many believe it’s time for a second look at what kind of additional applications, tools, standards or activities can benefit the model-driven Systems Engineering vision.

One area that is clearly missing in the current landscape is the means by which information from the variety of tools and applications used in Systems Engineering can be integrated. Given the breadth of the discipline it is not surprising that there is a considerable range of tools. To name a few, system engineering makes use of modeling and simulation tools, statistical analysis, reliability analysis, systems dynamics and life cycle planning and integration tools. Neither OMG SysML nor AP233 support a scenario that integrates data from those tools. Since data and model integration is a fundamental strength of RDF, one attractive approach to solving this challenge is to consider the application of Semantic Web technologies and modeling languages. This thought has led us to investigate the possibility and utility of a Systems Modeling Ontology expressed in OWL.
Why a Systems Modeling Ontology?
 
“IntegratioN”

Due to the distributed nature of Systems Engineering and the large, disparate datasets that SEs create, the use of Semantic Web technology solves many problems that other approaches find challenging.  Built-in global identifiers are just one example of the advantages that come from using Semantic Web standards. OWL and RDF provide a wide range of capabilities for modeling SE concepts and integrating data based on those concepts. SPARQL provides a powerful means for querying, mapping and constructing SE data.
By “Integration” we do not imply physically integration of SE data. Semantic Web tools are quite happy to pull data from almost anywhere on the Web and/or an Intranet. SE organizations can ‘publish’ their data to collaborating organizations enabling them to be queried as if they had been integrated. As illustrated in the following figure a Systems Modeling Ontology can be utilized as a Semantic Web standards-based means of integrating applications by mapping SE data through it
 Ontology-Based Interoperability for System Engineering
Interoperability between tools, in this case a SysML tool and another non-SysML tool, is achieved by mapping concepts and properties from a source model to a neutral model and then to a target model. The source and target models can be thought of as “Proxy Ontologies”.  An example of a proxy ontology is an OWL model for XMI. This is used to “lift” concepts and properties from the specifics of XMI into a pure OWL ontology of the source or target worlds.   The following figure shows some classes from the XMI Ontology.
Expressing XMI as an OWL Ontology
By representing concepts and properties in OWL, mappings and rules to and from the neutral model can readily be expressed. Mappings are done using OWL axioms and rules using a Model-Based Rules Execution (MBRE) approach based on SPARQL Inferencing Notation (SPIN).  SPIN is a simple extension to SPARQL that allows rules and constraints to be expressed on model resources. More details of SPIN can be found on the Web at http://www.spinRDF.org.

Inference

The systems produced using SE principles are often vast in scale and diverse in nature. Often they are “systems of systems” – a collection of task-oriented or dedicated systems that pool their resources and capabilities. Such an ‘emergent system’ offers more functionality and performance than simply the sum of the constituent systems.  Given this scale, systems engineers need tools to assist them in their analysis and verification of systems and systems of systems.
A Systems Modeling Ontology opens up the possibility of using Inference Engines to assist in that analysis. Inferencing also helps System Engineers understand the emergent properties of the systems of systems being analyzed. Because of their dynamic capabilities supporting data merging and query, Semantic Web technologies can enable real time analysis of systems after they are deployed. Mashups that add data about location, environment, operating condition, failure, etc. to data from the system design models themselves offer a potential for creating a whole new approach to systems operation.
Initial Work

An OWL ontology called SysMO, created by TopQuadrant, has been presented to INCOSE workgroups. This effort recently received renewed interest at the INCOSE International Workshop on “Model-Based Systems Engineering” at Mesa, Arizona in February, 2010.The workshop discussed the benefits of the pure approach to representing systems in SysMO as opposed to having to be constrained and affected by  UML constructs that result from OMG SysML being a ‘profile’ of UML.  The SysMO effort plans to take advantage of the well-developed Quantities Units Dimensions and Types (QUDT) ontology (see http://www.qudt.org) produced for the NASA Constellation program NExIOM (NASA Exploration Initiatives Ontology Models) project.  NExIOM formalizes the way machines (and people) refer to NASA Elements, their Scientific and Engineering disciplines, related work activities, and their interrelationships in the Enterprise. Through the use of agreed knowledge representations, information becomes intelligible and actionable to machines, tools, and people. Information can be found, associated, aggregated and reasoned over to generate products and inform decisions within and across diverse organizational groups.
For interoperability with the SysML world, concepts that are needed for representing SysML models are distinguished in the SysMO ontologies by having a unique ‘sysml’ namespace. For example, in the figure below, sysml:Block is from the SysML world. Some concepts from the SysML world are viable in the ‘pure’ SysMO model, for example, ‘sysmo:Connector’. The following figure shows some SysML concepts represented in the SysMO ontology.
Blocks and Ports Example from SysMO
The initial SysMO work has been influenced by and learned from the efforts of building extensive models of NASA vehicles and systems. A System is modeled using a “Structure, Behavior, Function, Interface (SBFI)” formalism. The NASA System Ontology extends SBFI with additional SE modeling constructs for perspectives (viewpoints) and aspects. While some SysML concepts are modeled directly, others are ‘freed’ from their UML dependencies. The modeling approach is modular with more general concepts in level 1 ontologies and more specific concepts in level 2, 3, etc. ontologies. The modeling approach also makes extensive use of OWL restrictions and Named Graphs. As an illustration of the modularity of the ontologies, the following figure shows how each named graph imports other named graphs. Note how the QUDT ontologies are involved in SysMO.
Ontology Architecture of the SysMO Ontologies 
Conclusion
The application of Semantic Web technologies is a natural next step in the evolution of the Systems Engineering IT frameworks and tools. Several organizations, including NASA, are already working in the area and can provide a good basis of experience.

Given recent events, it looks like 2010 will be seen as the year Systems Engineering and Ontology Engineering finally intersect in a serious way to produce what should be a very fruitful set of developments.