Information Innovation Office

The Information Innovation Office (I2O) is one of the six technical offices within DARPA, an agency of the U.S. Department of Defense that is responsible for the development of advanced technology for national security. I2O was created in 2010 by combining the Information Processing Techniques Office (IPTO) and the Transformational Convergence Technology Office (TCTO).^[1] The office focuses on basic and applied research in the areas of cyber security, data analytics]], and human-machine symbiosis.^[2][3]

Organization

The current I2O office director is John Launchbury, who joined DARPA as program manager in 2014 and was named director in 2015.^[4] Brian Pierce is the deputy director.^[5]

Research Areas

• Cyber: As human activity has moved into cyberspace, cyber threats against our information systems have grown in sophistication and number, and protecting and assuring information is a matter of national security. Progress in the cyber security of best-of-breed systems has been significant over the last few years, giving us hope that we are no longer facing an impossible task. So looking to the future, I2O challenges itself with the goal: Win at Cyber. The I2O defensive cyber R&D portfolio is focused on high-end cyber threats, including advanced persistent threats (cyber espionage and cyber sabotage) and other sophisticated threats to embedded computing systems, cyber-physical systems, enterprise information systems, and national critical infrastructure. I2O develops technologies that create software that is provably secure, applications that enhance cyberspace situational awareness, and systems for planning military operations in the cyber domain. Exploration of offensive methods is undertaken to inform the defensive cyber R&D and to establish viability of developed techniques with transition partners.

• Analytics: Exponential increases in computation, storage, and connectivity have combined over the past five years to fundamentally alter science, engineering, commerce, and national security. Going under names such as “big data,” “machine learning,” and “analytics,” empirical modeling and data-driven approaches are providing powerful insight and competitive advantage for astute practitioners from biology to sports to finance. Through new analytics, algorithms, and software ecosystems, the modern data-centric paradigm exploits the increasingly dense, detailed measurements produced by networked sensors to optimize products, services, operations, and strategy. I2O is working to keep the DoD at the forefront of data-driven design and decision-making with the goal: Understand the World. I2O explores fundamental mathematical and computational issues such as complexity and scalability and develops applications in high-impact areas such as intelligence, software engineering, and command and control. I2O coordinates its R&D with the national security community to ensure timely transition of tools and techniques.

• Symbiosis: The world is moving faster than humans can assimilate, understand, and act. At present we design machines to handle well-defined, high-volume or high-speed tasks, freeing humans to focus on complexity. I2O envisions a future in which machines are more than just tools that execute pre-programmed instructions: rather, machines will function more as colleagues. Towards this end, I2O sets a goal to Partner with Machines. The symbiosis portfolio develops technologies to enable machines to understand speech and extract information contained in diverse media, to learn, reason and apply knowledge gained through experience, and to respond intelligently to new and unforeseen events. Application areas in which machines will prove invaluable as partners include cyberspace operations, where highly-scripted, distributed cyber attacks have a speed, complexity, and scale that overwhelms human cyber defenders, intelligence analysis, to which machines can bring super-human objectivity, and command and control, where workloads, timelines and stress can exhaust human operators.^[6]

Active Programs

DARPA programs continually start and stop depending on national security needs and research results, and the high-turnover rate of program managers.^[7] The agency maintains an up-to-date list of active I2O programs on its official website. Some programs within I2O’s research areas are:

• Building Resource Adaptive Software Systems (BRASS): Creation of long-lived, survivable software systems that adapt to changing conditions on their own.^[8]
• Cleanslate design of Resilient, Adaptive, Secure Hosts (CRASH): Development of software techniques that allow a computer system to defend itself from hacks.^[9]
• Cyber Grand Challenge (CGC)
• Memex: Software that advances online search capabilities to extend into the deep web, the dark web, and nontraditional or multimedia content.
• Broad Operational Language Translation (BOLT): A program launched in 2011 to create automated translation and linguistic analysis.^[10]
• Explainable Artificial Intelligence (XAI): Create Artificial intelligence that can explain the decisions it makes.^[11]
• Plan X

References

1. DARPA: navigating the rapidly expanding infospace, SXSW, 8 March 2014. Retrieved 21 September 2016
2. How DARPA's I2O finds innovation on the edge, Federal Times, 2 October 2015. Retrieved 21 September 2016
3. Information Innovation Office Official Website, DARPA.mil, Retrieved 21 September 2016
4. Defense department taps Portland security expert to help make computer systems safer (Portland Business Journal, 9 June 2014. Retrieved 21 September 2016.
5. Brian Pierce: DARPA I2O Programs Seek Innovation through Info Ecosystem, Executive Gov, 5 October 2015.
6. http://www.darpa.mil/about-us/offices/i2o/more
7. Innovation at DARPA, DARPA Report, p.8. Retrieved 22 September 2016.
8. DARPA’s Researchers Plan Software That Wil Run for Hundreds of Years Without Upgrades, Slate.com, 13 April 2015. Retrieved 22 September 2016.
9. Reinventing the Internet to Make It Safer, New York Times, 2 December 2014. Retrieved 22 Septebmber 2016.
10. DARPA’s Newest Language Translator Would be Less Handheld Device, More Robot Assistant, Popular Science, 6 April 2011. Retrieved 22 September 2016.
11. DARPA wants artificial intelligence to explain conclusions and reasoning to humans, Next Big Future, 19 August 2016. Retrieved 22 September 2016.