Start your day with intelligence. Get The OODA Daily Pulse.
Editor’s Note: Chris bases this article on publicly available sources, most of which were produced by DIA. After reading we would appreciate inputs from any OODA Loop members on whether or not you think this is the right approach to creating a new data system. Seems like the government could be buying an existing solution that works instead of trying to code their own solution. Let us know what you think. -bg
The Defense Intelligence Agency (DIA) currently uses an old system called the Modernized Integrated Database (MIDB) to store and retrieve information on foreign intelligence – everything from Country order-of-battles (tanks, airplanes, etc.), infrastructures (runways, ports, etc.) to general foreign military capabilities. Problem is, it’s not so MODERN anymore. Originally built in the 1990’s, the MIDB uses manual processes to update and manage data. The proliferation of new sensors (intelligence gatherers) brings new data into DIA headquarters (and other production partners in the Services) at the speed of light. Analysts are then stuck trying to enter this wealth of information into a database that is neither agile nor automated.
Replacing MIDB is tricky since it serves many use cases and is connected to so many existing systems. The new data base must be trusted and accessible to hundreds of other applications (some modern, some legacy) that depend on this data. Each source of intelligence must be documented as authoritative sources. Each link must be tested for functionality and interoperability.
A new system is being coded from the ground up using contracted system integrators (non of which have ever done anything like this in the past). Machine-Assisted Analytic Rapid-Repository System (MARS) is the name of the new system.
To accomplish the transition, the MARS core is being built to operate along-side the MIDB until tested, validated and certified as trusted.
In December 2020, MARS was approved as a Major Program. To ensure MARS doesn’t stagnate like many large acquisition projects, they are using a modular approach which will let them tap into transformative ideas, technologies and capabilities from Industry throughout the lifecycle of the program. They have separated the prime contracts into four areas: System Integrator, Capability Developers, Data Transformation and Innovation. The Systems Integrator is responsible for integrating any capability modules that are purchased from other contracts. This approach will help keep the new MARS data “fabric” open to accept ideas and technologies as they are developed. This also keeps the door open for Industries to connect with DIA as they develop new ways (such as AI/ML) to perform database analytics, storage and retrieval.
Key to making sense of the mountains of data now being collected is having tools that are automated. MARS has the goal of being able to reach into many multiple data bases simultaneously, digging through thousands of intelligence categories, in seconds rather than hours. If the goals can be achieved via this approach, MARS will help the Analyst arrive at an assessment much faster and with confidence that all the key information sources have been included.