Desiging for Worst Case

Good designing is built around bad times

Early on my Substack journey I wrote a piece “The Days We Almost Died.” In it I talked about leadership and risk management failures underpinning a decision that left my team exposed to ISIS during a deployment to Syria. I am going to return to those days, not to re-litigate the command decisions, but to extract design and command lessons learned that I think apply now.

October Surprise

In 2017, my Special Forces Team (ODA) was supporting Syrian Democratic Forces working in southern Syria. For months, we moved and pushed our partners. We used mortars, artillery, HIMARS, and air strikes to support our partners during this fight. Our operations relied on good communications, knowing where our partners and where we were, and the ability for surveillance drones (ISR) and planes to see the ground.

As October rolled in, my team returned to the desert for a four-day push and ended up in a massive sandstorm that nearly cost us everything. The night before ISIS attacked the sandstorm rolled in. We moved locations to tuck behind some cliffs, hoping to mask ourselves and get out of some wind. The sand was so thick that while my team sergeant and I were talking about the next day we watched a Marine working with us nearly walk off cliff top while wearing his night-vision goggles. He never saw the edge. It was a moment of levity it what was an otherwise rough night.

The next morning, the wind picked up early and by 9 AM aircraft could no longer see the ground. Central Command grounded the ISR planes. Only jets could fly, but no matter how low they dropped they could not see the ground. Despite this, and that our operations were contengent on air support, we were still out firing mortars. Shortly after 9 AM, we were attacked by ISIS fighters that used the cover of the sandstorm to breakout of the river valley. This kicked off four days of close fighting that ended with me medically evacuating four soldiers and marines, closing our firebase, and retrograding (retreating) to reset and start our months of work again.

The sandstorm was not the issue. It exposed the weaknesses inherent in our decision making and planning. We were reliant on our technology, the ability to see the ground, communicate globally, track and know our position, and know where our forces were. In the heat of fighting, this was hard. The fog and friction of the sandstorm amplified those difficulties. The systems were not built for that level of stress. My ODA, which up until that morning had 30 radios for 12 people had four working radios, two on my truck, one on me, and one on my JTAC. We were left relying on runners and hand signals. Just as our communications faltered, one vehicle threatened to die¹, our partner locations tracking stopped, and everything digital became analog.

What didn’t change was our higher command’s desire for information. The commander was livid that he could not see the ground. He tried to interject, at one point attempting to fire HIMARS to support us. This was a panicked idea. It would have cleared our airspace, all planes would have to leave to avoid the flight path of the missiles, and taken away my ability to call on the jets to drop bombs through the sand. Command and control, taken for granted with ubiquitous technology was a drug and the command was instantly in withdrawals.

This was my two lessons. The first is companies must test their tech in brownouts. I don’t just mean in sandstorms. What happens when comms are completely jammed? What happens when the screen literally freezes on the tablet? If it is not designed and tested in those conditions, not lab conditions, soldiers will find it fails when they need it most. The second lesson was one for commanders. You have to train like every piece of technology will fail you cannot rely on any tech so much that it becomes the central pillar of an operation. For us, it was air cover and the ability to see. Losing that caused chaos.

Testing in these conditions is about graceful degradation. Does the tech let me know it’s not working? Is there a manual mode? If not, it’s unreliable, at best, at dangerous at worst. Passing these tests does not mean making a perfect device, it means making a device that performs its mission essential function 80% of the time in the combat conditions. Testing has to occur in combat conditions.

These lessons are not tactical. The U.S. military has placed more power in acquisitions leaders’ hands with rapid acquisition pathways and replaced Program Executive Officers with Program Acquisition Executives (PAE) and expanded their authority. These PAEs need to reconsider how they test tech. Everything must be tested in a degraded environment. That is where the 80% or better pass grade has to come from. 80% in ideal settings is worthless. Commands at every level have to know the limits of their tech and plans for when it fails.

1. Interestingly, this truck never started again under its own power. We stated it with a “slave cable” from that point forward and we retired it from use shortly after. So, we did have some luck that day.