The United States Department of Defense manages one of the largest procurement budgets in the world. Of the department’s $700 Billion annual budget, more than $200 billion is spent on goods and services from contractors.1 Purchases run the gamut from aircraft carriers to combat boots.
There are enormous challenges in procuring this array of goods and services with the right level of quality and speed while stewarding taxpayer dollars. One branch of the Department of Defense, the Defense Logistics Agency (DLA), is turning to digital models from XSB to reduce the cost, time, and errors associated with PDF-based procurement.
Defense Logistics Agency
The Defense Logistics Agency, the nation’s combat logistics support agency, manages the end-to-end global defense supply chain for the five military services. DLA procures more than $46 billion annually across several supply chains: subsistence (food/water), clothing and textiles, bulk petroleum and other energy products, construction material and equipment, personal demand items, medical material and equipment, and repair parts for land, sea and air systems.
DLA’s Clothing and Textile department manages more than $1.3 billion in demand for nearly 5000 Clothing and Individual Equipment items. Consider, for example, the Army Combat Uniform (ACU). The ACU is a complex garment incorporating camouflage, chemical treatments, fire retardant, NIR Signature management technology, and permanent IR IFF squares for identification with night goggles–just to name a few features designed to keep our soldiers safe and healthy.
The requirements to procure, manufacture and test the ACU are distributed across a web of related, but disconnected, technical documents including Specifications and Standards, Purchase Descriptions (PD), and other Technical Data Packages (TDP) with multiple Interim Changes. Product technical requirements are derived from documents authored inside (e.g., Military specs) and outside the Government (e.g., Non-Government Standards from ASTM, AATCC, and NFPA, etc.). These technical documents vary in age from thirty-year-old scanned PDFs, to more modern Microsoft Word and XML Documents. The purchase description for the ACU has references to 98 documents and 5053 pages from 13 document sources.
Different stakeholders including the military services, DLA Product and Contract Specialists, Government and industry testing labs, and suppliers to the government must find and use the requirements embedded in these documents. The documents are rich with technical information but were made for printing and reading. They are poor containers for technical data. They are not interoperable, and their uneven format makes it difficult for users to search, query, use, reuse, update, edit and manage. The problem is further complicated as different stakeholder groups establish multiple, independent collections of these documents on network drives, Microsoft SharePoint instances, on their local desktops, and even paper in folders.
Requirements in traditional documents are stored as static and disconnected objects, but they represent a dynamic web of concepts distributed across an ever-changing network managed by different authorities. The disconnected nature of a document-based approach makes tech data management difficult and can result in decisions based on inconsistent, incomplete, and out-of-date information. Different Clothing & Textile division stakeholders use the same information to perform different tasks during the product life cycle. A Product Specialist may support a military service by inserting contract-specific Interim Changes into an item requirement. In response, the manufacturer needs to alter a factory work instruction to ensure the finished product includes these changes. An industry test lab must change a First Article Test Plan to reflect the revised requirement. The DLA Product Test Center needs to know the impact of that change is reflected in manufacturer test reports. The problem with this disconnected document approach is that changes to one document are poorly communicated to other document stakeholders.
The Solution: TexSpecs
The DLA turned to XSB to help them create TexSpecs: “interoperable structured digital models of purchase descriptions, interim changes, and other specification-based technical documents which reduces the cost, time, and errors associated with PDF-based tech data management.”2
DLA’s transformation to a Model-Based Enterprise results in improved decision making and increased confidence that a design will perform as expected. This Model-Based approach enables users to query, edit, and analyze technical data that was previously locked in the document format. Linking the concepts within and between documents provides powerful change management and configuration control mechanisms. According to Deloitte, this approach can generate up to 65% process cost savings. It also helps DLA’s Clothing & Textile stakeholders avoid mistakes and rework resulting from decisions made with out-of-date information.
A Closer Look
XSB helped DLA convert documents from 13 different sources to interoperable, linked data models using Artificial Intelligence and semantic technology. The collection of these former documents, now models, is stored as a knowledge base, or graph, and is called the Digital Model Library (DML). The knowledge graph establishes a single, enduring, authoritative source of truth that captures the state, history, and relationships between tech data sources. Changes made to a document model in the DML propagate throughout all affected data and systems, ensuring stakeholders have accurate and up-to-date information. This Model-Based digital thread is the product DNA for the management of the associated clothing and individual equipment items. Digital models of Clothing & Textile documents provide many advantages over PDF documents:
- consistent and instantaneous integration of Interim Changes
- enhanced configuration management across stakeholders,
- Automated and rapid analysis to ensure compliance with government procurement procedures and interoperability with other digital models from Government and Industry.
In addition, digital models can be exported as MS Word documents enabling the Agency to retain its traditional, document-based processes and workflows while undergoing digital transformation.
The knowledge graph, or DML, can be accessed by humans through a Web browser, or by machines and systems through a powerful Application Programming Interface (API). The DML now contains 146 Purchase Descriptions, 620 Government Specifications, Standards, and Commercial Item Descriptions, and 2130 Non-Government Standards from ASTM, AATCC, and others.
2 DLA C&T Modernization Efforts with DLA MUST Program TexSpecs& SRP Tool Brief for Joint Advanced Planning Brief to Industry (JAPBI)