Many of the technological tools the Defense Logistics Agency depends on today, including artificial intelligence, wouldn’t exist if not for databases. Even before high-speed processors and probability-based algorithms, databases organized information in retrievable formats, a necessary precursor for AI agents returning in-depth answers to user prompts.
Information has long needed to be arranged, but databases, which require automated data processing, are relatively new. One of the first organizations to integrate multiple databases was DLA, formed as the Defense Supply Agency in 1961, about the time “database” entered the English language.
Database pooling with DIDS
DSA became a database manager when it inherited the federal catalog in November 1961. At the time, the catalog consisted of 38 million 5-by-8-inch cards, 7 million of which the agency trucked from Washington, D.C., to Battle Creek, Michigan, the headquarters of its new Defense Logistics Services Center. Activities throughout the federal government stored the remaining 31 million. Digitizing this massive archive in DSA’s 80-character format permitted national stock number information to be integrated with other logistics data. DSA called the program governing these interactions the Defense Integrated Data System.
DIDS took a decade to deploy. Managers running its databases received information in various ways, including over the phone and by mail, but most entries arrived by wire transmissions. Military units used specially designed machines to transmit punch-card data to DLSC mainframes over telephone lines. The Defense Communications Agency operated the lines and DSA’s Defense Automatic Addressing System routed the transmissions.
DIDS stored data on magnetic tape capable of holding 13.5 billion characters. While a government auditor chastised the agency for creating a program that was onerous to build and unable to keep up with expanding requirements, DIDS nonetheless consolidated existing databases, decreased the time it took to screen new catalog items and interfaced with material management programs. As revealed in the auditor’s report, what prevented DIDS from being truly transformative wasn’t nearsighted vision or a deficient design but the limited computing capabilities of its two Burroughs 6700 and one IBM 360/65J computers.
DIDS covered more than cataloging. The system also helped DLSC with repurposing military equipment. While matching excess in one service to the needs of another depended on those services uploading on-hand quantities, the databases they populated belonged to DIDS. Another database mission was the Integrated Disposal Management System, established by DSA and operated by its Defense Property Disposal Service, a predecessor of today’s DLA Disposition Services. When property disposal employees updated this system, DLSC’s mainframes updated other DIDS databases. Supply condition codes were updated, as well.
Special database missions
Not all database pooling involved DIDS. Automatic data processing equipment was scarce in the mid-1960s, with only 2,600 computers in the federal government. At the request of the Defense Department, DSA established a program in 1964 that screened this equipment for potential reuse. While covering multiple inputs and two owner types, the databases feeding the program didn’t relate to cataloging and weren’t integrated into DIDS.
The Defense Industrial Plant Equipment Center also had a database mission. DIPEC was a primary-level field activity DSA established in 1962 to manage manufacturing equipment. The center had multiple components, including the agency’s only maintenance activity and a repository that could loan machines. This last mission depended on databases outside DIDS tracking where equipment was and where it could go.
Another one of DSA’s database missions resulted from the Vietnam War. In 1968, the Defense Department asked the agency to design an automated program for placing veterans in federal jobs. The Vietnam Era Veterans Employment Referral System consisted of two information flows in constant need of updating, one for veterans and one for job openings. Run out of the Defense Electronics Supply Center in Dayton, Ohio, VEVERP targeted 225 occupations and found jobs for former service members in the Defense Department, Department of Agriculture, Post Office and Civil Service Commission.
Material management programs
Not all database missions helped DSA procure items and logistics services. While reutilization databases and the Integrated Disposal Management System were incorporated into DIDS, DIPEC inventories and VEVERP were administrative programs unrelated to the agency’s core mission. DSA leadership wanted databases in set formats so they could feed programs automating operations. DIDS, for example, was incorporated into the Standard Automated Materiel Management System, DSA’s first platform digitizing logistics functions. Subsequent database-dependent management programs included the Mechanization of Warehousing and Shipping Procedures and Defense Fuels Automated Management System.
After DSA became the Defense Logistics Agency in 1977, it began sharing data outside the organization. The need for cooperation became evident during the First Gulf War when poor shipping container management led to overordering and delayed cargo. The agency and U.S. Transportation Command addressed this problem by pooling databases. A quarter century later, changes to the Federal Acquisition Regulation mandated another logistics partner — corporations — register with the agency’s Commercial and Government Entity program. CAGE is a longstanding catalog database that had been competing with a proprietary system. A current digital advancement project called Technical Data Management Transformation integrates databases compiled by project executive offices and original equipment manufacturers into language acquisition professionals can use to compete contracts.
Database management in the current technological environment
Today, DLA combines in-transit visibility, contractor information and repair parts data in a central repository, or data lake. This lake is a modern version of DIDS. The difference is technology. Developments outside the defense arena have allowed data in noncongruent formats to be pulled, assimilated and converted into media humans can comprehend.
While the most advanced form of pulling, assembling and computing data is AI capable of decision-making and independent action, earlier iterations have been flattening DLA procurement for decades. The standardization that began with SAMMS and continued with Enterprise Business Systems can be seen in the agency’s approach to aviation consumables. DSA rejected Defense Department suggestions that it manage aircraft parts early in its history due to items in the relevant federal supply classes being complex, technical, subject to engineering changes and requiring reparable programing. With today’s management systems blurring differences in ordering procedures, the agency has been able to combine its aviation, land and maritime parts management into one supply chain, DLA Weapons Support. This centralization wouldn’t have happened without database consolidation.
The same pooling concept is also helping DLA digitize its supply chains. In the near future, databases from EBS and other systems will be aggregated and their data assigned tags. These tags convert the agency’s data lake into a data fabric, allowing supply chain, finance, property and human resources programs to pull information from one location.
Supply chain mergers and digitized supply chains cannot occur without consolidation. As with artificial intelligence, DLA’s digital journey started with the humble database.