Federal R&D Requirements for Big Data Investment

Back in mid-September, the federal government’s Networking and Information Technology Research and Development Program (NITRD) released its long-awaited Supplement to the President’s Fiscal Year 2020 Budget. The document outlines the new fiscal year’s agenda for federal agencies leveraging emerging technologies for research and development purposes. My colleague, Christine Fritsch, published a summary of the NITRD Supplement’s contents, focusing particularly on artificial intelligence, which can be found here. Today’s post details big data-related requirements from six of the NITRD’s eleven Program Component Areas (PCAs): Artificial Intelligence R&D, Computing-Enabled Human Interaction, Communications, and Augmentation (CHuman); Computing-Enabled Networked Physical Systems (CNPS); Enabling R&D for High-Capability Computing Systems (EHCS); High-Capability Computing Infrastructure and Applications (HCIA); and Large Scale Data Management and Analysis (LSDMA). Some artificial intelligence requirements are detailed below as well, providing industry with an understanding of the types of work that federal agencies could fund in support of their big data/AI R&D efforts. See the NITRD supplement linked above for more information on which agencies have requested budgets for each area.

Artificial Intelligence R&D: Algorithm/software development, platforms for human–AI collaboration, improved visualization and AI-human interfaces, enhanced AI security, shared public datasets and environments for AI training and testing, AI performance benchmarking and standards development.

CHuman: Improved interfaces between humans and intelligent systems (e.g., robots, intelligent agents, autonomous vehicles and machine learning systems). Modeling and simulation and workforce training.

CNPS: Modeling and analysis tools; development of interoperability standards; smart technology/Internet of Things R&D; autonomy research; workforce development and training.

EHCS: Exascale computing systems; quantum, neuromorphic and probabilistic computing R&D; advanced computing prototypes; benchmarking of new computing architecture; new modeling, simulation and analytics tools; workforce development.

HCIA: Acquisition of high-capability computing systems; algorithm/software development; development of high-capability computing infrastructure and ecosystems; computing testbeds; modeling/simulation tools; collaborative work environment development for high-capability simulation and data analytics; expansion of workforce.

LSDMA: Development and testing of decision-making analytics; data interoperability and scaling R&D; real-time analytics; high-speed data transport capabilities; translating R&D into operational tools.

Areas of Opportunity

Educating the federal workforce about big data-related technologies, even in the scientific community, runs through most of these programs, indicating an opportunity for industry partners to do consulting work.

Software development is another service mentioned multiple times, particularly as that service pertains to advanced analytics development. Agencies have been responding to the need for software development expertise in two ways. First, they have been leveraging cloud-based production environments, with spending on these rising from $29M in fiscal 2016 to $118M in fiscal 2018. Expect this trend to continue. Second, many have adopted agile development methodologies. There is still room for growth in this space.

Modeling and simulation services should be in demand. This will be especially true at the Department of Defense, which Section 957 of the fiscal 2017 National Defense Authorization Act required to make greater use of.

Developing human-machine capabilities is also common on the list, with work needed particularly across the DOD, but also at Civilian sector agencies like the National Aeronautics and Space Administration.