Hacking for Defense: Week 3
SWJ Note: See Hacking for Defense (H4D) @ Stanford – Week 1 for an update on week 1 and an overview of Stanford University’s Hacking for Defense course and Hacking for Defense: Week 2 for an update on the second week.
We just held our third week of the Hacking for Defense class. This week the 8 teams spoke to 108 beneficiaries (users, program mangers, etc.), we held a Customer Discovery workshop, we started streaming the class live to DOD/IC sponsors and other educators, our advanced lecture was on Product/Market fit for the DOD/IC and we watched as the students solved their customer discovery obstacles and started getting closer to their customers.
(This post is a continuation of the series. See all the H4D posts here. Because of the embedded presentations this post is best viewed on the website.)
Customer Discovery in the DOD/IC Workshop
We normally hold a Customer Discovery workshop during the evening the first week of the class. But spring break and the “How to Work with the DOD” workshop got in the way. So we inserted an abbreviated version at the front of this week’s class.
When working with the DOD/IC there are some unique obstacles of “getting out of the building and talking to customers”. For example, members of the DOD will not respond to ”cold calls” and those in the Intel community won’t even tell you their names. In addition, most of the sponsors are working on classified problems. So how do teams understand the customer when the customer can’t tell you what they do? The Workshop talked about how to address those and other Discovery issues.
See presentation here.
Team Presentations: Week 3
After the Customer Discovery workshop the 8 teams presented what their hypotheses, experiments and what they learned outside the building this week.
Team Right to Boom (previously named Live Tactical Threat Toolkit) is trying to help foreign military explosive ordnance disposal (EOD) teams better accomplish their mission. The team originally was developing tech-centric tools for foreign teams to consult with their American counterparts in real time to disarm IED’s, and to document key information about what they have found. Now they are honing in providing accurate high-volume post-incident IED reporting.
Last week this team was floundering. They had confused getting interviews and building minimal viable products with truly trying to “become the customer.” We strongly suggested that there was no way that could understand the day in the life of an explosive ordnance disposal expert by just listening to them – they needed to stand in their shoes. So to their immense credit the team suited up in full bomb disposal gear and got of the building. They earned our respect (and a name change for the team.)
See the Right to Boom video here (turn up the volume!).
See the Right to Boom presentation here.
Team Fishreel is combating “Catfishing”; where someone is impersonating a specific person, or a person with a specific interest or an organization or an entire service. The team is working to develop a technique to score how likely it is that a given online persona is who they claim to be, and how that conclusion was reached.
This week, the teams learned that you don’t need be coding to get customer feedback. Instead, they built wireframe in PowerPoint and got useful feedback from their sponsor. They validated their main hypothesis with users and validated that the current process for verifying identities of potential catfishers is “extremely labor intensive”. They found that their MVP is tracking toward a solution that users could find very helpful.
They refined their understanding of Spearfishing vs. catfishing finding that Spearfishing is the number 1 source of malware, cyber attacks, and network threats.
See the Team Fishreel presentation here.
Team Capella Space is launching a constellation of synthetic aperture radar satellites into space to provide real-time radar imaging.
This week the team learned a ton. They mapped out competitive offerings, found that Government funding is not the proper channel for Capella, but did find that the Coast Guard is currently in dire need of situational awareness at high resolution and that military customers want access to raw data; commercial customers highly value processed data for actionable insights.
See the Team Capella Space presentation here.
Team aquaLink is working to give Navy divers a way to monitor their own physiological conditions while underwater (core temperature, maximum dive pressure, blood pressure and pulse.) Knowing all of this would give divers early warning of hypothermia or the bends.
This week they validated that divers will want real-time alerts regarding vitals (and put up with the additional gear/procedures) of issues that threaten mission success. The found that navy medical researchers want data on vitals, the rebreather (air consumption), and the dive computer (dive profile). Their hypotheses going forward are that a heads up display is the ideal form of information transmission during a dive and system should be modular to allow for the integration of evolving technology (geolocation and communication).
See the Team aquaLink presentation here.
Team Guardian is working to protect soldiers from cheap, off-the-shelf commercial drones. What happens when adversaries learn how to weaponize drones with bullets, explosives, or chemical weapons?
Guardians current hypotheses is that they have to provide drone detection, identification and protection against attacks from drones or swarm of drones. And that the user will be a 19 solider not trained to use complex equipment.
See the Team Guardian presentation here.
Team Narrative Mind is trying to understand, disrupt, and counter adversaries’ use of social media. Current tools do not provide users with a way to understand the meaning within adversary social media content and there is no automated process to disrupt, counter and shape the narrative.
The team is coalescing around the idea that the two minimal viable products for their sponsor are, 1) automatically generate an organizational chart of a target terrorist groups over time, and 2) generate a social network map of how terrorist groups interact with each other.
See the Team Narrative Mind presentation here.
Team Sentinel initially started by trying to use low cost sensors to monitor surface ships in a A2/AD environment.
The team has found that their mission value is really to enable more efficient and informed strategic decisions by filling in intelligence gap about surface ships in an A2/AD environment via:
- Increased number of data streams (i.e. incorporate open source data)
- Automated data aggregation (i.e. from disparate sources) and analysis
- Enhanced intel through contextualization
- Improved UI/UX
See the Sentinel presentation here.
Team Skynet is also using drones to provide ground troops situational awareness (Almost the inverse of Team Guardian).
The team invalidated the hypotheses that military/commercial systems exist that could already solve the problem. In addition, they originally believed that soldiers on foot needed a deployable drone system. They discovered that drones are best used with teams with vehicles or for short ranged dismounted operations.
See the Team Skynet presentation here.
Advanced Lecture: Product/Market fit in the DOD/IC
The advanced lecture for week 3 was on the unique needs of finding Product/Market fit in the DOD/IC. Pete Newell described why a solution in the DOD fails and then described “battlefield calculus” – how two identical sounding missions (and their inherent problems) are actually radically different based on what echelon of force executes them, by the size of force, their location, even by how well they are trained. Despite the obvious, people still try to deliver “one-size-fits-all” solutions. To properly insure a solution is actually used it is important to become familiar with the pattern of life of the user and their unit.
Pete also pointed out that teams need to “Look for Conflict” between what may have been provided to solve a similar problem and the solution the teams are about to recommend. You needed to ask: Are the circumstances similar? Or are there a myriad of conditions present that will invalidate what was a good solution under different circumstances?
See the presentation here.
Mission Model and Value Proposition Canvas
To students, “who are the beneficiaries?” feels fuzzy on day one. And given most of them had no exposure to the DOD or Intel Community it’s not a surprise. The reason we have the teams talk to 10-15 people every week, is that with enough data they can begin to fill in the details. A few of our guests have commented how knowledgeable the teams were in talking about the sponsor organizations and problems.
That said, listening to the team presentations there was a wide difference between teams in how well they understood that the definition of “beneficiaries.” Many of teams were still listing names of organizations rather than the title and archetype of the people who mattered/cared/decided/users, etc.
Understanding who the beneficiaries are is critical to understanding the rest of the mission model canvas.
When the students have a more nuanced understanding who are the individual beneficiaries is when they can build a detailed Value Proposition Canvas for each beneficiary that makes sense. (Several teams had Value Proposition Canvas of organizations, some had fewer Proposition Canvas than they had beneficiaries, some Proposition Canvases were so generic it was clear that had insufficient data on individual needs of specific archetypes, etc.)
This is all par for the course and part of the student learning. We now need to sharpen their focus.
An after class action for the teaching team is to read through every team’s week 3 presentation slide-by-slide and give each team a detailed, written, box-by-box critique of the right-side of their Mission Model and Value Proposition Canvas. We want to help them get this right.
Sponsor Education – a Network Begins to Form
The teaching team, liaisons, mentors and DIUx are all working their networks to get students relevant beneficiaries to talk to. (More about what a wonderful asset DIUx has been in a future post.) Joe and Pete are continuing to work hard on educating the sponsors about their role. (We are collecting all our learning in an Educators Guide so other universities can run the class.)
One emerging unexpected benefit, is that Pete and Joe are continuing to expand the network of innovators in the DOD/IC who are helping our student teams. I’ve had several critique our presentations and offer suggestions on the nuanced parts of the IC mission and acquisition system I didn’t understand.
Live Streaming the Class
The DOD/IC sponsors who gave us these problems were curious about how the teams were learning so rapidly. (Others in their commands and agencies wanted to watch as well.) So this week we began to live-stream the student presentations. And other universities who want to offer this class have begun to have their educators watch the class. (We’ll be offering a train-the-trainer educators class later this year.)
Lessons Learned from Week 3
- Teams still running at full speed.
-
Understanding beneficiaries is critical to understanding the rest of the mission model canvas.
- Written team-by-team offline critique is needed to keep them on course.
-
Support is coming from lots of places in the DOD/IC.
- DIUx and our liasons have been great in connecting the students.