In a blog post published last April, I outlined how the Defense Advanced Research Projects Agency intended to spend more than $200M in fiscal 2016 on Research, Development, Test, & Evaluation programs with a big data-related component. The qualifier “big data-related” meant R&D projects that employed algorithms, modeling, advanced analytics, and/or visualization techniques as part of the overall work. The programs were categorized into those intended to develop big data technology and approaches as a “primary” requirement versus using big data technology to reach another goal. An example of big data as a primary requirement would be a project that had as its goal the development of a big data algorithm or analytic. An example of big data as a secondary requirement would be the use of an analytical approach to develop a more rigorous cyber defensive tool. The cyber tool is the primary requirement in this latter case, while the big data analytics used to develop it are secondary. As a refresher, here is the table from that April 2015 blog post which showed those programs.
Thanks to the release of the Department of Defense’s FY 2017 budget request, the data can be brought forward another year. Doing this reveals the following evolution of big data-related spending at DARPA.
DARPA projects that its planned investment related to big data will rise from approximately $216M in FY 2015 to nearly $269M in FY 2017, a planned increase of 24.5%. The difference in totals from the previous year’s post and this year’s has largely to do with my identifying additional programs that have a big data element.
Concerning the programs themselves, two with big data-related work (i.e., ENGAGE and Nexus 7) which received RDT&E dollars in the FY 2016 budget have dropped out in FY 2017. Conversely, several additional programs have been identified for FY 2017. Among these is “Big Mechanism,” an effort to automate computational intelligence for biology, cyber, economics, social science, and military intelligence. In FY 2016, part of the work being performed will include developing causal models that relate cancer phenotypes to genotypes using biological big data.
Other newly identified programs harness medical/biological big data too, including Neuro-Adaptive Technology and the Analysis and Adaptation of Human Resilience. Neuro-Adaptive Technology didn’t appear in last year’s list because the big data element of the program got overlooked. I’ve included it this year because the program’s objective is to leverage visualization and modeling techniques to map neurological functions. The Analysis and Adaptation of Human Resilience program, by contrast, is a new start in FY 2016 with $13M slated for exploring new methods of maintaining and optimizing warfighter health in response to emerging infectious diseases. The big data portion of this work involves using “sophisticated algorithms to identify important patterns of survival” among animal species.
Wrapping up, DARPA’s investment in big data related R&D will continue to be strong in FY 2017, providing business opportunities for companies that do work on complex systems, the integration of sensor data, the development of advanced algorithms, and for applying big data analysis to biological systems.