Manufacturing Systems Integration Program

Supply Chain

Supply Chain Integration

Annual FTEs: 11 NIST FTEs

10 Guest Researcher FTEs

21 Total FTEs

Challenge

The manufacturing sector will continue to provide the core of our nation’s real wealth creation in the foreseeable future. Innovation and competition within this sector will take place at the level of supply chains or “value chains[3],” however, not at the level of individual companies. These value chains will evolve into a network of global, collaborative partnerships. These partnerships will develop rapidly when new market or technology opportunities appear. They will dissolve just as quickly when those opportunities disappear.

As part of this evolution, manufacturing will become less resource-intensive and more knowledge-intensive. This shift will require a new capability – the ability for all partners to exchange and assimilate information instantaneously. The challenge for the Supply Chain Integration program is to foster and promote that capability by developing and demonstrating an infrastructure for the testing and integration of automated systems that exchange information throughout the supply chain.

Overview

The Supply Chain Integration program has produced a number of syntax- and quality-based testing and validation tools that are now available as NIST Web services or as free stand-alone tools. To address new challenges, we have begun shifting our focus to content-based testing[4]. This is necessary to ensure that the semantics of the information being exchanged is correctly understood. In addition, the program is pursuing new approaches to manufacturing standards development that facilitate, rather than hinder, the automation process. The foundation for this infrastructure comes from the completed Automated Methods for Integrating Systems (AMIS) project, which provided a fundamental basis for our information exchange and application integration.

The AMIS project defined a general approach for automating the exchange of information semantics between any two software applications. This approach was demonstrated using a “Request-for-Quote” supply-chain scenario and standards from the Open Applications Group (OAG) and the Chemical Industry Data eXchange (CIDX).

The Inventory Visibility and Interoperability (IV&I) project used the AMIS approach to demonstrate automated exchange of inventory-on-hand data between global automotive suppliers. Working with the Automotive Industry Action Group (AIAG) and with European and Korean collaborators, MSID developed numerous semantic technologies, which provided the initial set of components for our integration infrastructure.

The Materials Off Shore Sourcing (MOSS) project began last year. It is adapting the IV&I testing infrastructure to include business processes[5]. MOSS aims to reduce the uncertainty in door-to-door shipping times by simplifying and integrating the information flow among the dozens of players involved. To date, MSID has developed a conceptual data model for all the information objects, a model for all of the messages, and a number of testing tools.

Key Accomplishments and Impacts:

• Developed syntax and quality-based tools that are routinely used by developers from both government and industry, resulting in a 30% reduction in the time to develop new e-commerce specifications. AIAG has projected more than $200M savings from suppliers using interoperable IV&I
  applications.
• Potential savings have not been documented for MOSS, but the impacts of poor interoperability are known. They include a substantial increase in premium shipping costs, more than 20 days of excess inventory maintained, 80% of all information re-keyed with an average cost of $20 each, and more than 40 days difference between best (21 days) and worst (63 days) shipment times from Europe. If successful, the MOSS project will have a significant impact on all of these numbers.

Future Directions and Plans:

• Enhance existing infrastructure to facilitate the exchange of manufacturing and business data and provide support for testing and validating manufacturing and business specifications. Continue the shift in testing focus from syntax-based to content-based testing. This is necessary to ensure that the content of the information being exchanged is correctly understood.
• Pursue new approaches to manufacturing standards development that will promote/ease, rather than hinder, the automation process. Launch the Virtual Supplier Network project that will enable OEMs (e.g., GM, Ford, DaimlerChrysler) and suppliers to match requirements with capabilities automatically over the Internet through the addition to the MOSS infrastructure of components that will address two important complications: the fact that the applications involved are not known in advance, and the influence of business negotiations on the exchange of technical data.

Awards and Recognition:

Board Membership

Projects

Supply Chain Integration

Introduction

Supply chain integration enables supply-chain partners to exchange information automatically. This exchange takes place between disparate software applications that need to work together to satisfy a unified manufacturing or business need. These applications likely run on multiple, geographically dispersed computer systems and platforms, and were generally not designed with integration in mind. These and other issues make integration projects complicated and extremely expensive.

Studies have shown that most integration efforts fail to achieve their goals. Successful integration involves four stages: (1) specification creation, (2) specification validation, (3) development of conformance and interoperability test methods, tools, and data sets, and (4) implementation. Each of these stages builds upon results of the previous stages, so problems in any stage can greatly increase the likelihood of overall failure.

MSID researchers have identified a number of such problems, including:

• Inaccurate capture and/or documentation of data interchange rules and constraints – Unreliable manual processes result in a time-consuming and error-prone cycle of interpreting and refining data processing rules and constraints.
• Inconsistent specification design quality – Often caused through lack of consistent practices, even while using formal modeling languages.
• Inadequate content validation – Primarily due to lack of appropriate content validation tools.
• Informal and unstructured specification definition – Leads to ambiguity and misinterpretation.
• Syntax-only integration standards approach – Leading analysts have estimated that 35% to 65% of system integration costs are due to semantic issues that syntax-only standards cannot address.
• Local management of data interchange rules – Results in conflicting standards.
• Non-adaptable, non-extensible implementations – Leads to the proliferation of incompatible dialects.

The Supply Chain Integration program is developing reference models and practical toolkits that will address these issues in innovative ways. Together, the program’s products will make possible accurate specification creation, formal validation and testing, and rapid application integration.

Strategy

The program has two main thrusts: the development of tools for the testing and validation of existing supply-chain interface standards; and the development of an infrastructure to automate application integration.

Figure 1. Conceptual View of the Supply-Chain Integration Stack [D]

The first thrust is based on the standards life-cycle activity model of standards, which we presented at our last meeting in 2005. That model addresses the creation, use, and maintenance of any standard. We have used this model to frame our research and the development of tools to support validation and testing of an existing standard (see Figure 1). In the following section, we provide overviews of these tools primarily in the context of the W3C XML Schema, because of its widespread adoption by industry.

The second thrust is based on the approach developed in the AMIS (Automated Methods for Integrating Systems) project, which we also presented in preliminary form in our last meeting in 2005. The AMIS strategy is to derive from the published interface specifications for a software system (application program interfaces, messages, exchange files, protocols, specifications, and documentation) an understanding of the roles in the business processes the system was built to support. That understanding must be captured in formal models that contain definitions of the business entities, properties, processes and rules in a machine-readable form suitable for automated reasoning.

The AMIS project produced a high-level description of the major components of a semantics-based integration infrastructure for supply chains (see Figure 2). That infrastructure contains local ontologies, a reference ontology[6], mappings, translators, and connector transformations. The project also demonstrated how to build those components for a ‘request-for-quote’ supply-chain scenario and related standards from the automotive and chemical industries.

Figure 2. Conceptual View of Application Integration [D]

Lessons/ideas from AMIS became a crucial part of AIAG’s Inventory Visibility and Interoperability (IV&I) project, which sought to develop an agreed-upon set of e-Kanban[7] messages that would allow suppliers to communicate inventory data to the rest of the supply chain. MSID’s role was to collaborate with European and Korean partners on a semantics validation project designed to show that the automotive industry could lower cost, improve speed to market, and reuse IT investments by using emerging semantic technologies to define and implement messages among suppliers.

This project was important for four reasons. First, it applied the AMIS approach to a real integration problem using real commercial IV&I applications. Second, it demonstrated the potential of using semantic technologies to automate the integration process. Third, it showed that it may be possible to separate that process into two parts: a private part that only the users know about and a public part containing models, tools, and agents that do all the real work. Fourth, it showed the potential to further separate those models, tools, and agents into one collection that is specific to a certain branch of manufacturing and another collection that is generic.

AIAG’s Material Off-Shore Sourcing (MOSS) project is an initiative designed to improve business procedures and information drivers controlling the intercontinental shipment of goods. AIAG studies assert that improvements in the accuracy of information conveyed—and agreement in how it is to be interpreted—will result in tangible reductions in overall supply-chain transit times and measurable decreases in the variation of transit times. A direct result will be reductions in buffer-stock inventory and expedited/premium transportation expenditures.

In FY07, MSID began working with MOSS participants from government, AIAG, and the vendor community to develop the necessary recommendations and standards, test associated Electronic Data Interchange (EDI)[8] messages, develop a testing infrastructure, and conduct demonstrations using an ocean-shipment scenario. We will again use the AMIS approach and extend the IV&I infrastructure to evaluate a real business process.

Supply Chain Integration

AMIS (Automated Methods for Integrating Systems)
(Status: complete in 2006)

Challenge/Problem Addressed:

Supply chain integration, even when it is based on existing syntactic interface standards, is an error-prone, costly, time-consuming, repeated, human-intensive engineering activity.

Objective(s):

• Reduce the effort required to exchange supply chain semantics by devising models, methods, algorithms, and tools to automate engineering activities that are now done by human systems integrators.

Accomplishments:

• Developed a reference architecture that describes the major components of a semantics-based integration infrastructure for supply chains. That infrastructure contains local ontologies, a joint reference ontology, semantic mappings and translators, and connector transformations.
• Demonstrated the reference architecture based on (1) a request-for-quote supply chain scenario, and (2) ontologies based on two different supply chain standards: one from the automotive industry and one from the chemical industry, and, (3) mappers and translators to go from one standard ontology to the other.

Customers and Collaborators:

• OAGi (Open Applications Group, Inc.)
• CIDX (Chemical Industry Data eXchange consortium)

Supply Chain Integration

ATHENA/IV&I (Inventory Visibility and Interoperability)
(Status: complete in 2007)

Challenge/Problem Addressed:

Integrating off-shore suppliers into an existing supply chain so that inventory levels are visible to all partners is costing billions of dollars a year within the transportation sector.

Objective:

• Develop and implement an integration infrastructure, based on the AMIS reference architecture, to exchange inventory information automatically between automotive supply-chain partners.

Accomplishments:

• Developed a monitoring tool to help test and validate supply chain integration implementations that use the Business Process Specification Schema (BPSS) and Collaboration Protocol Agreement (CPA) specifications. The tool checks whether each message has the right sender and receiver, message sequencing, and time constraints. The tool is implemented as a Java applet, which enables it to run in web browsers.

• Industry and government partners have adopted the Content Checker tool to aid the consistent application and validation of supply chain transaction specifications in real manufacturing transactions. This tool precisely specifies and extends conformance testing based on the semantics defined in a transaction schema, or content standard. Currently, the Content Checker works with transaction specifications based on the XML Schema language.
• Developed a collection of ontological models, semantic annotation tools, semantic reconciliation tools, and semantic translation tools based on existing OAG inventory standards and developed jointly with U.S, Korean, and European partners.

Customers and Collaborators:

• Automotive Industry Action Group
• SAP
• Open Applications Group, Inc
• KORBIT[9]

Supply Chain Integration

Materials Off-Shore Supply Chains (MOSS) (Status: complete in 2009)

Challenge/Problem Addressed:

The complexity of information transfers adds substantial logistics delays forcing OEMs to use premium shipping and warehouse extra inventory, which adds billions of dollars a year in shipping costs.

Objective(s):

• Propose recommendations and standards
• Define associated EDI messages
• Develop an integration infrastructure
• Conduct demonstrations to show the automated exchange of logistics information across the entire logistics business process

Accomplishments:

• Developed a MOSS conceptual model of the objects and relationships in ocean-freight transport and messaging supporting the management of ocean-freight manufacturing supply chains.
• Defined the MOSS message type structures, which include the ~100 key properties for the management of ocean-freight manufacturing supply chains, and mappings of this information into EDIFACT[10] messages.
• Developed tools to assess the conformance of MOSS participants’ messages to the MOSS recommendation. The current tools check for correct syntax formation and the presence of required information
  elements.
• Designed a message metamodel for representing the structure of an EDI- or XML-based message type. Its use in MOSS may demonstrate a strategy for decoupling concerns of message syntax from the task of message processing.
• Built a Queries/Views/Transformations (QVT) mapping engine, which is the major component of the conformance validation tooling. It will be used to map information from EDI messages to a form consistent with the MOSS conceptual model.

Planned Future Accomplishments:

• Proof-of-concept conformance testing demonstration
• Detailed business process and interoperability demonstration

Customers and Collaborators:

• Automotive Industry Action Group
• Honda of America Manufacturing
• U.S. Customs and Border Protection
• Bosch
• Daimler Chrysler Corporation
• Ford Motor Corporation
• General Motors Corporation
• Global Commerce Systems

Supply Chain Integration

Integration Standards Testing Tools
(Status: complete in 2009)

Challenge/Problem Addressed:

Studies have shown that most integration efforts fail to achieve their goals. Poorly designed interface specifications and inadequate test methods and data are still major causes of these failures.

Objective(s):

• Provide industry with a suite of open tools and test methods that allow quick and easy assessment of XML-schema based interface standards.

Accomplishments:

• The Naming Assister tool is used by standards development organizations to evaluate a specification’s consistent use of naming. The tool maps terms used to assemble element names or type names against a table of allowable terms. It can also check the construction of compound names against the International Organization for Standardization’s ISO 11179 recommended naming convention.
• Extended the use of the Schema Validation tool that allows users to validate transaction schemas and transaction instances against the W3C standard specification for XML schemas.
• Extended the Schematron Editor tool to provide a Java-based GUI tool for business analysts to create, view and modify
  Schematron files easily.
  
  The tool includes a number of wizards to (1) facilitate specification of constraints and more precise semantic definitions of business-content standards, and (2) test an XML instance file against the constraints defined by the current Schematron file.

Planned Future Accomplishments:

• Enhance instance validation tool that allows users to validate specific instance data content with either their own uploaded schemas, or from publicly available schemas including OASIS UBL v1.0, Grants.gov v1.0, and OAGIS v9.0.
• Extend the Quality of Design (QOD) tool as a more sophisticated tool suite, introducing Web 2.0 capabilities that allows users to define customized schema testing profiles based upon specific tests defined by them or collected from tests previously defined by others.

Customers and Collaborators:

• Open Applications Group Inc. (OAGi)
• Automotive Industry Action Group (AIAG)
• FIATECH[11]
• Un/CEFACT
• Navy
• IRS
• GSA
• U.S. Air Force
• OCIO Council

Supply Chain Integration

Supply Chain Center
(Status: complete in 2011)

Challenge/Problem Addressed

Supply chain challenges span all industrial sectors, but the solutions are typically addressed within the context of a given sector, even at NIST, resulting in duplication and inconsistency.

Objective(s):

• Create a NIST-wide interdisciplinary
  Supply Chain Center where multiple
  laboratories can come together to share results and enhance each other’s work

Accomplishments:

• Developed a web-based collection point both for sharing efforts across NIST, and to provide one-stop shopping for customers looking for NIST work relevant to supply chain interoperability. All of the tools mentioned above can be accessed through this web site.

Planned Future Accomplishments:

• Enhance and evolve the web-based collection content and presentation
• Expand the user base of the web-based supply chain center

Customers and Collaborators:

Other NIST Laboratories, including:

• BFRL
• EEEL
• ITL