Real-Time Factory Information Project

Effectively moving information along the production and assembly processes within a factory is a key aspect of smart manufacturing, but is inefficient due to the lack of standard information models. The inefficiency is manifested as the extra time and cost needed to translate between proprietary equipment, and information lost when converting into simplistic data formats. The benefits of optimal process planning and scheduling can only be partially realized in the face of the barriers put up by this wasted effort and information deficiency. Standard information models are needed to fully represent the information needed to program smart factory equipment. This project will develop the measurement science necessary to develop, validate and test these standards, enabling the seamless flow of information along the manufacturing and assembly equipment that make up a smart factory.

Develop and deploy measurement science and standards to program smart manufacturing equipment on the factory floor, eliminating the wasted time and effort needed to achieve optimal performance at each stage of production, delivering results to standards bodies by 2014.

The technical idea is to develop and test new integrated information models for programming manufacturing and quality measurement equipment. The manufacturing industry has been challenged to "establish consistent, efficient data methods for all industries^[1]" to realize the smart factory, with actions to conduct gap analyses, develop new data methods, protocols and standards, and conduct beta tests. This project responds to the call for national laboratories to participate in analysis and validation, through two measurement science elements: the development of conformance tests for factory equipment and the applications that program them, and validation testing in open public pilot tests that measure the effectiveness of the standards in meeting industry needs. One key research element of the project is determining a core set of information that can be reused among the many manufacturing processes possible, a task that will be shared with the domain experts from the industry groups with whom the project will participate. Information modeling today is primarily done using the Extensible Markup Language (XML) Schema format for defining syntax, which is supported by a wide variety of tools and large experience base. The measurement science element on which NIST staff will focus is the development of conformance testing software based on XML Schema, with a delivery of these tools to accredited standards organizations as a primary output. Although useful for catching a large proportion of errors, these software tools are only able to measure conformance to syntax, but not how well the systems understood the meaning of the information. To measure how well equipment and programming software handles the meaning (semantics), pilot tests will be done in which automated regression testing takes place on combinations of equipment, and involves human verification according to precise definitions of correct behavior that is hard to measure automatically. A key measurement science element for this task is the determination of a reduced set of tests that adequately cover equipment function from the combinatoric space possible.

The problem addressed by this project is that incompatible system components impede the smart factory. The objective is to increase compatibility through measurement science support for standards and protocols, which is needed to enable effective and efficient information exchange. Expected outcomes of this project will follow a general timeline for each standard starting with standards gap analysis, harmonization, use cases, a standards architecture, new standards versions, validated standards, early implementations and pilot tests, conformance tests, conforming implementations, and certified implementations. The research plan is crafted to provide research support that is broadly valuable throughout manufacturing (i.e., discrete, assembly, process, or batch).

A "information models and validation tests" research element will help ensure the proper definition of each standard: what information is modeled and how that information is modeled. This research task will include gap analysis, harmonization studies, interface architectures, use cases, schema architectures, on-machine quality, semantic conformance studies, the product manufacturing information (PMI) required for assembly optimization, and harmonization of the QIF (quality information framework) family of standards with standards-based PMI (product manufacturing information) and quality control plans.

A "verification tests" research element will allow manufacturers to ensure that implementer claims to conformance are quantitatively verifiable. This research task will require the application of formal conformance methods, such as design of experiments, legacy standards conformance test maintenance, semantic conformance, conformance criteria definition, baseline standards necessary for standards-based closed loop and smart manufacturing, and all new conformance suites with automatic code generation for small & medium sized corporations.

The "open public pilot tests" research element will allow the entire standards community to verify and expand on use cases and conformance classes early in the standards development phase. This research task will require early (pre-adoption) pilots at individual manufacturer sites and web-based conforming code development uploading tool.

The research conducted will be highly collaborative between project researchers, manufacturers, technology providers, academics, and other NIST research colleagues (such as ITL's Software and Systems Division, EL's Systems Integration Division, and EL's Applied Economics Office), due to the multi-disciplinary nature of the research.

• Outcome: ISO 10303-238:2007, STEP-NC (Numerical Control)
• Outcome: ANSI and ISO versions of DMIS, v4.0 and v5.2
• Outcome: Conformance suites for all DMIS (Dimensional Measuring Interface Standard) 5.x versions, consisting of conformance tests, accompanying files, user documentation, and test files
• Outcome: Conformance suites for I++ DME (Inspection plus-plus Dimensional Measurement Equipment) v1.4 and v1.5, consisting of conformance tests, accompanying files, user documentation, and test files
• Output: The OMAC (Organization for Machine Automation and Control) Memorandum of Agreement between MTConnect Institute and the Dimensional Metrology Standards Consortium (DMSC) for standards-based on-machine quality;
• Impact: Siemens PLM Software's NX CMM Inspection Programming application was DMIS certified in September 2010.

Expected standards-related outcomes: In the QIF family of standards: QMResults, QMPlans, and QMStatistics versions (in priority order); new versions of DMIS and I++ DME; MTConnect and QMResults integration; vendor implementations of STEP-NC.

Expected impacts: Continued substantial benefits to important U.S. corporations through use of DMIS (Dimensional Measuring Interface Standard) and I++DME (Inspection plus-plus Dimensional Measurement Equipment), which enable CMM (coordinate measuring machine) programming and a standard interface to a CMM; increased DMIS certified software; MTConnect machine tool interface has enjoyed broad manufacturer and vendor support and enables dashboards for scrap detection, power monitoring, and overall machine tool effectiveness and increased implementations and benefits are expected; increased CAM software vendor implementation of STEP-NC and benefits to all manufacturers.

[1] Smart Manufacturing Leadership Coalition, Implementing 21^st Century Smart Manufacturing Workshop Summary Report, June 24, 2011, p. 16.