On Tuesday, I listened to eight teams of Stanford students present their solutions to current national security problems on the final day of H4D: Hacking for Defense. They tackled Twitter analysis, hostile drone identification, ship tracking, spotting illegal activities through cloud cover around the globe, and other defense challenges.
The class, the first of its kind, was created by a group of Stanford faculty members and advised by former U.S. Secretary of Defense William Perry. It had support from multiple branches of the U.S. government, including the National Security Agency, the Army, the Navy, the Air Force, the Army Cyber Command, the Veterans Administration, and other agencies that offered up a list of problems of interest.
Team Fishreel’s Travis Noll (left to right), Chris Gomes, Maya Israni, and Gabbi Fisher think the tool they developed to help determine who or what is behind a Twitter account will have commercial as well as military applications.Photo: Tekla Perry
The eight four-person teams were a mix of students ranging from undergraduate freshmen to PhD candidates, with academic interests including business, engineering, and computer science. Each group had 10 weeks to define a real-world problem, propose a solution, run its solution by at least a hundred potential end users and military “customers” (that is, the people who would eventually approve funding for deployment), demonstrate at least a minimally viable prototype (in the vernacular of the class, an MVP), and figure out how that solution could be deployed.
All the teams got somewhere—at least moving the needle towards a better solution—even if they didn’t completely solve a problem. Many teams intend to continue developing their technologies; one team even scooped up $200,000 in angel funding to support the next phase of its project.
The problems and solutions:
- Team Aqualink aims to give Navy Seals operating underwater better geolocation data for themselves and their submersibles by developing a small, releasable GPS buoy.
- Team Capella Space aims to launch Synthetic Aperture Radar (SAR) small cube satellites that can better see through cloud cover. This would make it possible to monitor illegal activities in the equatorial region, including drug trafficking, illegal fishing, and piracy. (This is the team that has already secured angel investment. They expect to launch satellites next September.)
- Team Narrative Mind is building automated tools to detect emerging narratives in the constantly churning ocean of 140-character Twitter messages. This, the team says, will enhance cyberdefense by providing forecasts of U.S. adversaries’ plans.
- Team Fishreel is working on Web-based tools that analyze Twitter names to give detailed summaries of Twitter behavior, and suggest whether the user is genuine, a human imposter, or a bot. The tools are also capable of identifying similar accounts that may be part of an organized propaganda campaign. Though the team originally designed the tools for national security analysts, they see a lot of commercial uses too.
- Team Sentinel, working with the Navy’s Seventh Fleet, aims to provide better real-time information about where ships are. Today’s tools often give incorrect or contradictory location information.
- Team Skynet is combining real-time video feeds from drones with existing maps in order to automatically identify humans and geolocate them in one fell drone swoop.
- Team Right of Boom is working to improve an app already in use for tracking IED discoveries. Now that the class has ended, the team will be consulting for the Joint Improvised Threat-Defeat Agency.
- Team Guardian is developing a tool that can take video from a friendly drone feed, use it to distinguish drones from birds, and then identify and classify enemy drones in the area. This classification extends to determining what type of aircraft the enemy drones are and what they might be carrying, and suggesting the best countermeasures.
The class, besides helping the defense agencies with nagging difficulties, was designed to teach students entrepreneurship and the Lean Startup methodology.
In conversations with the students after their presentations, it seemed that the key common revelation was finding out how important it is to talk to the eventual users before building a solution. Indeed, almost every team started out with a vastly different approach to solving a problem than the one with which they eventually ended up. At some point during the 10 weeks, each made the classic Silicon Valley “pivot.”
Team Aqualink, for example, set out to develop technology to monitor the vital signs of Navy Seals, before discovering that getting good location information was a more pressing problem. In both cases, Team Aqualink’s approach would require new hardware. But they were the exception.
Team Sentinel pivoted from hardware to software.Photo: Tekla Perry
More typical pivots took development plans from hardware to software; once the teams got a handle on the problems they were trying to solve, they realized the hardware was out there, but better software was needed in order to make it useful. Team Sentinel, for example, initially thought that providing better information about ship locations for the Navy would require building a new sensor network. They discovered that there were plenty of sensors out there; the real need was for better data aggregation, analysis, and visualization. Team Skynet thought giving soldiers in the field better information about the threats around them would require some kind of heads-up system; existing pad computers, it turned out, are a fine interface if humans in a drone image can be automatically identified. Team Right of Boom also was looking to augmented reality glasses and remote expertise to help those attempting to defuse IEDs—that is, until they were told by a potential user, “The last thing I’d want while defusing a bomb is a voice in my ear telling me what to do.” They pivoted, focusing their efforts on improving an existing app for the collection and analysis of IED incidents.
The Stanford faculty members who developed this class—Steve Blank, Tom Byers, Joe Felter, Pete Newell, and Jackie Space—say they want to roll it out to at least 50 other U.S. universities in the next three years. Their aim is to turn it into a technology-based version of the Reserve Offficers Training Corps, or ROTC—a new way for university students to contribute to the country. Wrote Blank in the course announcement:
Today if college students want to give back to their country they think of Teach for America, the Peace Corps, or Americorps. Few consider opportunities to make the world safer with the Department of Defense, Intelligence Community and other government agencies. The Hacking for Defense class will promote engagement between students and the military and provide a hands-on opportunity to solve real national security problems.
“By creating a national network of colleges and universities,” Blank says, “the Hacking for Defense program can scale to provide hundreds of solutions to critical national security problems every year.”
It’s off to a strong start.
Tekla S. Perry is a senior editor at IEEE Spectrum. Based in Palo Alto, Calif., she's been covering the people, companies, and technology that make Silicon Valley a special place for more than 40 years. An IEEE member, she holds a bachelor's degree in journalism from Michigan State University.