Z.01: Data in a Crisis
Unlike the initial run of Downrange Data posts, this one is not based on my anecdotal experience. I was in Okinawa when we evacuated Afghanistan. I left North Carolina for Australia before Hurricane Helene ripped through this fall. What do these two events have to do with data? More than you’d think.
Both were crises, where time was short, and a lot of coordinated action needed to happen very quickly. Both also saw bottom-up data solutions that were key to resolving the crisis.
Our stories about past crises often look for individual leaders who took the reins and solved problems through decisive action. We seem to always want a ‘great men of history’ narrative. But in reality, our killer app as humans has always been how we can cooperate across large groups. Data literacy is in many ways just the ability to do this at ever greater speed.
‘Let me just put the spreadsheet in there and share it…’
Coordination at speed was exactly the problem facing the airmen trying to keep HKIA open in August of 2021.1 As the coalition tried desperately to evacuate Kabul amidst a surging Taliban, managing the single runway left to them became perilous. More than just ensuring a landing plane doesn’t impact one taking off, the airmen had to manage the ‘slot times’ of aircraft on the tarmac. HKIA didn’t have the plethora of parallel runways and aprons you’d find at the Dallas-Fort Worth or Atlanta airports. Too many planes landing could effectively choke off everyone’s ability to get back out.
At the start of the crisis, the airmen of the 332nd Expeditionary Operations Support Squadron were doing all of the coordination of flights via phone call. Very rapidly the number of planes needing to be coordinated outstripped the minutes left in the day. Capt. Adam Solomon remembers his frustration with the old process. ‘I was like “this is absolutely ridiculous, I cannot be on the phone 24/7”.’ I can certainly empathize.2
So, the airmen pivoted, standing up a Google Sheet and quickly sharing out permissions to other services and even other governments. ‘I personally never talked to the president or any other heads of state or the Prime Minister of the United Kingdom, but I gave permission to their staff and that’s how they were briefed with all that information,’ Solomon said.
On top of the challenges of flight ops, just getting people to HKIA presented hurdles. As city after city outside Kabul fell, thousands began to desperately try to flee the country. This included not only international civilians and NGO workers, but also Afghans who had worked alongside NATO forces for two decades in support of the now floundering Afghan government. The military mission in Afghanistan was doing all it could, but a myriad of volunteers, many retired veterans of Afghanistan, jumped online to help.
A mix of volunteer groups and defense contractors stepped into the gap, including Task Force Dunkirk, Team America, and Quiet Professionals. They ‘…helped coordinate air assets, ground force assets, gate access, document processing, food and shelter and biometrics screening.’ And they did it with data. Quiet Professionals, a firm heavily populated with retied SOF operators, had recently acquired multiple data firms, to include Echo Analytics Group, and Scion Analytics. Firms which specialize in data, machine learning, and artificial intelligence.
These groups were able to build data trackers where Afghans could register their documents. They also pushed location data to the NATO mission in Afghanistan on those isolated and unable to get to HKIA. While the SOF veterans in these organizations understood operations, the key ingredients were ‘tech-savvy’ 30-somethings who could quickly code solutions to capture, collate, share, and see data. They were the ones that took those initial Google Sheets and Microsoft Excel band-aids and began coding new systems.
Pineland Needs You
Hurricane Helene wasn’t the man-made disaster that was the collapse of the Government of Afghanistan. But it shared many of the same problems, all ripe for data driven solutions. The devastation also hit very close to home.
The damage left in the wake of Hurricane Helene needed disparate kinds of help, and like HKIA, there were a lot of people who were prepared to build bottom-up data-literate solutions. In the days after the storm, an old special forces friend of mine, long a leader in the SFCT, reached out to me.3
The SFCT mobilized to help the families impacted by the flooding and destruction of Hurricane Helene, particularly those families in North Carolina. This was in no small part because every SF soldier owes the families of rural North Carolina a debt. Our culminating exercise, Robin Sage, is supported by local volunteers across 50,000 square miles of North Carolina. Every operator who earned a Green Beret since 1974 did so fighting in ‘Pineland’, a made-up country peopled by role-players drawn from the community in square in Helene’s track.
I wasn’t much help unfortunately, but I could see in the group chat bottom-up data solutions were already being shared. There was a need to catalogue and coordinate the flow of logistics, separate disinformation from real reports of families in need, and a way to share and see the latest imagery from commercial satellites. To coordinate the group has already setup their own ATAK server, a tool many of us first met when calling in close air support in Afghanistan.4 New members to the chat were prompted to download ARCGis and were given a video tutorial to show them how to get started. One brave soul was even trying to get everyone to jump from Signal to a better suited service like Slack.
Civilian volunteers weren’t the only ones applying data to Helene’s recovery efforts. When the 18th Airborne Corps was given permission to join the effort, they brought with them their MSS — Maven Smart System. MSS was built for managing indirect fires. So why was it now helping FEMA and the DoD map roads and cell towers impacted by the storm, as well as coordinate logistics with isolated communities in need?
MSS stands out as an example of a different way for the army to build software. The reason it could be pivoted from field artillery to FEMA is in part because of an atypical design process which focused on how data was gathered and moved. It shrank the distance between end-user and the engineers to effectively zero. And instead of just coding software that met years old requirements specifications, MSS was built to be adapted and updated constantly as new data streams became available. MSS was built with crisis in mind.
At the same time, MSS stands out as an indictment of our leadership, because it’s how we should have been coding over a decade ago. Part of the reason we have so many bottom-up solutions to our data challenges is because our top down imposed one are so horrible. Well before the crisis occurred, leaders failed, leaving their subordinates to improvise solutions with the digital duct tape they had available to them.
MSS shows what we can make when ‘visionary front-line leaders’ are supported by ‘senior leaders being willing to champion a program’. That was the finding of Emelia Probasco, a senior fellow at Georgetown University’s Center for Security and Emerging Technology (CSET), in her recently published report on the development of MSS. John Amble sat down with Probasco and Igor Mikolic-Torreira, the director at CSET in the latest Modern War Institute podcast. Give it a listen, because that’s what I want to talk about next.
Hamid Karzai International Airport, now renamed back to Kabul International Airport.
Special Forces Charitable Trust is a non-profit organization that provides resilience programs, resources, and assistance to Special Forces Soldiers and their families.