Daniel sent us this one, and I love it because it's one of those questions where the answer is simultaneously way more boring and way more terrifying than you'd expect. He's asking: how do logisticians plan for environments where resupply is intermittent or impossible? Think Antarctic research stations, military outposts, naval vessels at sea. What's the actual planning process, the inventory management system, and who are the people doing this work? There's a phrase that kept coming to mind when I was thinking about this — "logistics of last resort." Because when the next delivery window is months away and the cost of running out is catastrophic, everything changes.
The cost of running out in Antarctica isn't "we'll grab more from the warehouse." The cost of running out is "people die." And that's not hyperbole. The British Antarctic Survey requires overwintering staff at some stations to have their appendix removed preemptively before deployment. You cannot be evacuated from the South Pole between February and October. The last flight leaves, and that's it. You're in a closed loop.
"Closed loop" — that's the phrase. And it's what makes this fundamentally different from a Walmart or an Amazon fulfillment center. Those systems are designed around flow. Stuff comes in, stuff goes out, algorithms reorder. These extreme environments are designed around a finite bubble of resources with a hard deadline for the next infusion. Once that window closes, you are living on what you've got.
That's the framework we should use to think about this. There are really three archetypal closed-loop environments. One: Antarctic research stations, where resupply happens during a roughly four-month summer window from October to February. Two: naval vessels at sea, particularly submarines and aircraft carriers, which carry everything they need for deployments of ninety to a hundred and twenty days. Three: remote military outposts — forward operating bases in places like Afghanistan or Arctic early-warning stations — where resupply might be intermittent due to weather, terrain, or hostile activity.
The resupply calculus is completely different in each of these. But they share one thing: the planning horizon isn't "what do we need this week?" It's "what do we need for the next six to nine months, calculated down to the individual tube of toothpaste?
Let's start with Antarctica, because it's the most extreme version of this. The US Antarctic Program operates three year-round stations: McMurdo, South Pole, and Palmer. McMurdo is the logistics hub. During the summer season — roughly October through February — McMurdo receives approximately eight million gallons of fuel and eight million pounds of cargo. That's the entire annual resupply for the continent's US operations. Everything for the winter-over crew has to be on the ice before the last C-17 or LC-130 flight leaves, typically in late February.
Eight million gallons of fuel and eight million pounds of cargo. In four months. That's not a supply chain, that's a firehose.
Here's where it gets interesting from a planning perspective. The winter-over population at McMurdo is typically a hundred and fifty to two hundred people. At the South Pole station, it's about fifty. You have to calculate, for every single item those people might need, the consumption rate per person per day, multiply by the number of person-days you're covering, and then add a buffer.
Let's get concrete. How do you calculate how much toilet paper a hundred people need for six months?
You start with a consumption rate. Military and Antarctic logistics planners use standardized estimates for literally everything. Toilet paper might be estimated at, say, one roll per person per week. A hundred people times twenty-six weeks is two thousand six hundred rolls. Then you add your buffer — typically fifteen to twenty percent — so you're ordering about three thousand rolls. That's your number. And someone has to justify that number in a cargo manifest.
The buffer isn't "just in case people use more toilet paper than expected." The buffer is for breakage, spoilage, contamination, or the cargo pallet falling into a crevasse on the traverse from McMurdo to the South Pole.
Which brings up the last mile problem. The South Pole station is eight hundred miles inland from McMurdo. Getting supplies there involves either LC-130 ski-equipped aircraft doing multiple flights, or the South Pole Traverse — a ground convoy of tractors pulling sleds across the ice sheet. That traverse takes about thirty to forty days each way. So your resupply planning for the South Pole has to account for the fact that the stuff you ordered in October might not arrive until December, and if anything's wrong, you have a very narrow window to fix it.
Who's actually doing this planning? Is it some AI running simulations? Is it a room full of generals?
It's a surprisingly small team of humans with spreadsheets. At McMurdo, the key person is the Cargo Supervisor, supported by a team of maybe three to five logistics specialists. They use a system called the Cargo Management System — CMS — to track every pallet, every container, every item. But here's the thing that surprised me: many Antarctic supply officers still use paper cargo manifests as backup. When the power goes out or the satellite link drops, you still need to know what's in which container.
There's something deeply satisfying about that.
The US Antarctic Program's logistics are actually managed through a joint operation between the National Science Foundation and the US military, called Operation Deep Freeze. Every year, the military provides airlift and sealift support. A cargo ship arrives at McMurdo in January with the heavy stuff — fuel, construction materials, vehicles. Flights bring in the priority cargo. And everything is prioritized using a tiered system. Medical supplies and fuel are what they call "mission critical" — those get top priority and the highest buffer margins. Fresh food is a morale item. It gets cut first when weight or space is tight.
The lettuce is the canary in the coal mine of Antarctic logistics.
And there was a great example of this in twenty twenty-one. A miscommunication in the cargo manifest led to McMurdo running out of fresh eggs for three months during the winter-over. Three months with no fresh eggs. The workaround was powdered eggs, which apparently are... fine, but nobody's happy about it. The point is, a single line item error in a spreadsheet in October meant a hundred and fifty people were eating powdered eggs in July.
That's the fragility here. These systems look robust because they're planned so meticulously, but they're actually incredibly brittle. One mistake, one mislabeled pallet, and you're eating powdered eggs for a quarter of a year.
That's Antarctica, where at least the resupply window is predictable. Let's talk about submarines, where the constraints are even tighter. A US Navy fast-attack submarine carries enough food for about ninety to a hundred and twenty days for a crew of a hundred and thirty to a hundred and fifty sailors. The total food load weighs roughly forty thousand pounds. Every single pound matters because the submarine has to maintain precise trim and stability. You can't just throw extra cases of Gatorade on board.
That's the detail that gets me. The weight of the food affects the submarine's ability to stay level underwater.
And that means the provisioning is planned to the last ounce. The Navy uses what's called a "load list" system. Every meal for the entire deployment is pre-planned down to the ingredient level. They operate on a menu cycle — typically a twenty-one-day rotation of breakfast, lunch, and dinner menus. The load list specifies exactly how many pounds of chicken breast, how many gallons of milk, how many cans of green beans are needed to execute that cycle for the full deployment.
It's not "order what we think we'll need and adjust." It's "here is exactly what every sailor will eat on day forty-seven of the deployment, and we've bought precisely that amount.
The US Navy has a manual for this — it's called the S-9, the submarine food provisioning manual. It specifies two point five pounds of food per person per day, broken down into specific categories. There's a formula. And the Supply Officer — typically a Lieutenant Commander — is responsible for executing this with a team of culinary specialists. But here's the key difference from Antarctica: on a submarine, the crew can "pull" items from the load list, meaning the cooks can decide which specific meal to prepare from the pre-planned menu. But nobody can reorder. There is no Amazon Prime for submarines.
That "push versus pull" distinction is important. In Antarctica, supplies are pushed based on forecasted needs — the planners in Denver or Christchurch decide what gets sent. On a submarine, there's pull capability within the pre-loaded inventory, but no ability to change the total quantity or add new items. Different failure modes. Antarctica's failure mode is "we forgot to order enough of X." A submarine's failure mode is "we have X, but it all spoiled because the cold chain failed.
Let's talk about the cold chain, because it's one of the most underappreciated challenges in all of this. On a US Navy aircraft carrier — which by the way serves eighteen thousand meals per day, over one point six million meals over a ninety-day deployment — perishable food is managed across five different temperature zones. Frozen, deep chill, standard refrigeration, cool dry storage, and ambient. Each zone has specific humidity and temperature requirements. If a compressor fails, you can lose tens of thousands of dollars of food in hours.
A carrier has a Supply Department of over two hundred sailors. Two hundred people just managing inventory and food service. That's larger than most companies' entire logistics operations.
The Supply Officer on a carrier is a Lieutenant Commander or Commander — a mid-career officer with years of specialized training. They're managing what is essentially a floating city of five thousand people with a warehouse that can't be restocked until the next port call or underway replenishment. And they're doing it with a taxonomy that most civilians have never heard of: the Class I through IX supply classification system.
Break this down. What's Class I through IX?
This is the system the Department of Defense formalized in nineteen fifty, and it's still used by every US military branch today. Class I is food and rations. Class II is clothing and individual equipment. Class III is petroleum, oil, and lubricants — POL in military speak. Class IV is construction materials. Class V is ammunition and explosives. Class VI is personal demand items — think soap, cigarettes, snacks from the ship's store. Class VII is major end items like vehicles and weapon systems. Class VIII is medical supplies. Class IX is repair parts and components.
When a logistics officer is planning for a deployment, they're not thinking "what stuff do we need?" They're thinking "what's our Class I requirement for ninety days, what's our Class VIII formulary, what's our Class IX mean time between failure forecast?
And each class has its own planning methodology. Class I is consumption-based — meals per person per day. Class VIII is formulary-based — what medical conditions are we likely to encounter, what drugs and supplies treat those, what's the shelf life? Class IX is the hardest. That's spare parts, and it's forecast using mean time between failure data — MTBF. You look at every pump, every valve, every circuit board on the ship, you know its historical failure rate, and you stock spares accordingly.
It's not "order extra of everything." That's the misconception most people have. "They just buy more than they need.Space and weight are severely constrained on a ship, on a submarine, on an aircraft flying to the South Pole. Every item has to be justified. Over-ordering isn't just wasteful — it creates disposal problems, it takes up space that could be used for something else, and in the case of a submarine, it literally affects the vessel's ability to operate.
That's where the buffer percentage becomes a strategic decision, not just a safety margin. Fifteen to twenty percent is standard for consumables. But for critical spares where a failure would mean a mission abort — the part that keeps the reactor cooling pump running, for instance — you might carry a hundred percent redundancy. Two of everything. And you still might get unlucky.
Let's talk about what happens when things go wrong. Because all this planning — the spreadsheets, the cargo manifests, the S-9 manual — it's all a best guess about a future that doesn't cooperate.
There was an incident in twenty twenty-three. A US Navy carrier had to air-drop insulin to a submarine. A diabetic crew member's supply had been miscalculated. Someone, somewhere, made an error in the consumption rate calculation, or the buffer wasn't adequate, or the sailor's dosage changed mid-deployment. And suddenly you have a submarine somewhere in the Pacific with a medical emergency that can't wait for a port call. So a helicopter flies out from a carrier with a cooler of insulin, and they do an at-sea transfer.
That's the human factor. All the systems in the world, and it still comes down to someone filling out a spreadsheet correctly. And the person filling out that spreadsheet on a forward operating base in Afghanistan might not be a logistics specialist. It might be an E-6 or E-7 — a staff sergeant or gunnery sergeant — who got handed the additional duty of supply officer and a clipboard.
That's the reality in remote military outposts. At a large base like Bagram, you'd have a full logistics battalion. But at a small forward operating base with fifty or a hundred soldiers, the supply chain is managed by a senior NCO who probably has other primary responsibilities. They're doing inventory counts by hand, radioing in resupply requests, and making judgment calls about what's critical and what can wait.
They're operating in a "push" system where higher headquarters is deciding what gets sent based on standard load lists. The staff sergeant at the FOB can request things, but there's no guarantee they'll arrive. Weather, enemy activity, higher-priority missions — any of these can delay or cancel a resupply convoy or helicopter sling load.
This creates a fundamentally different psychology around inventory. In a normal supply chain, you want to minimize inventory because it ties up capital. In an extreme environment, inventory is survival. Nobody at the South Pole is worried about carrying costs. They're worried about what happens if the traverse breaks down and the next resupply is delayed by a month. So the buffer isn't a cost — it's insurance. And the premium is paid in cargo weight and storage space.
Let's talk about the consumables versus durables split, because that's where a lot of the planning complexity lives. Consumables are straightforward — food, fuel, medicine, toilet paper. You calculate the burn rate, multiply by time, add buffer, done. Durables — vehicles, generators, scientific equipment, aircraft — require a completely different approach.
With durables, you're not forecasting consumption. You're forecasting failure. And that's where the MTBF data comes in. For every component on a generator, you know how many hours of operation it typically lasts before failing. You know your planned operating hours for the deployment. You stock spares for the components that are most likely to fail within that window. And you accept that for some low-probability failures, you just won't have the part.
In Antarctica, if that part breaks and you don't have a spare, you're not getting one for six months. So you improvise. The history of Antarctic exploration is basically a history of improvised repairs. Duct tape, baling wire, and desperation.
The US Antarctic Program maintains what's called a "winter-over survival stock" — a cache of critical supplies that is never touched except in genuine emergencies. It's the logistics equivalent of a reserve parachute. You plan not to need it, but if you do need it, it's the only thing between you and catastrophe.
How do you plan for a medical emergency when you have a finite formulary? The doctor at McMurdo or on a submarine has a specific list of drugs and supplies. If someone develops a condition that requires a drug not in the formulary, what happens?
That's the nightmare scenario. In Antarctica, during the winter-over, there is no evacuation. The South Pole station is colder than the surface of Mars for much of the winter. Aircraft can't operate. The hydraulic fluid freezes. So the station doctor — usually a single physician — has to manage whatever comes up with what's on hand. They carry a broader formulary than a typical primary care practice specifically because they can't refer out. But there are limits. If you need an MRI or surgery beyond basic procedures, you're out of luck.
That's why the British Antarctic Survey requires the preemptive appendectomy. It's not paranoia. It's math. The probability of appendicitis in a young adult population times the number of person-days times the consequence of no surgical capability equals an unacceptable risk. So you remove the appendix before deployment.
That same calculus applies to every item in the cargo manifest. What's the probability we'll need this? What's the consequence of not having it? What's the weight and volume cost of carrying it? It's a continuous triage process.
This brings us to the "what if" scenarios. How do you plan for a supply ship that's delayed by a month? Or a traverse that hits a crevasse field and has to reroute? Or a compressor failure that takes out your cold storage?
You plan in layers. The first layer is the buffer — fifteen to twenty percent extra on consumables. The second layer is substitution — if fresh produce runs out, you have frozen and canned. If eggs run out, you have powdered. The third layer is rationing — if a resupply is delayed, you cut non-essential consumption. The fourth layer is the emergency stock — the reserve parachute. And if all of those fail, you're into the fifth layer, which is "figure it out or die.
That's not hyperbole either. The history of polar exploration is littered with expeditions that got the fifth layer wrong. Shackleton's Endurance expedition is the famous example — they lost their ship, but Shackleton's leadership and planning meant everyone survived. The Franklin expedition got it wrong, and everyone died.
Modern Antarctic logistics is designed to make sure nobody ever has to test that fifth layer. And it's remarkably successful. The US Antarctic Program has been operating continuously since nineteen fifty-six without a single winter-over fatality due to supply failure. That's an extraordinary safety record.
Let's shift to the practical takeaways, because most of our listeners are not provisioning a South Pole station. But the principles here are surprisingly applicable.
The first one is the buffer percentage concept. Whether you're stocking a pantry, a workshop, or a small business, the fifteen to twenty percent redundancy rule is a good heuristic. For items where running out would be catastrophic — critical medications, backup power, emergency water — go higher. For items where running out is merely annoying — snacks, entertainment — go lower or zero. But make it a conscious decision, not an accident.
The second takeaway is the consumption rate method. Track what you actually use per person per day or per week for a few months. Not what you think you use — what you actually use. That baseline is the foundation of any good inventory system. Most households have no idea how much toilet paper they use in a month. A submarine supply officer knows to the roll.
The third takeaway is the push versus pull tradeoff. This is a strategic decision that applies to any supply chain. Are you going to forecast demand and pre-position supplies — the push model — or are you going to let actual consumption drive reordering — the pull model? In extreme environments, push dominates because pull is impossible. But in everyday life, most of us use a hybrid. Your pantry is push — you stock it based on what you think you'll need. Your weekly grocery run is pull — you buy what you've run out of.
The insight from extreme logistics is that you should be explicit about which model you're using for which items, and you should know the failure mode of each. Push fails when the forecast is wrong. Pull fails when the resupply is delayed. Different risks, different mitigations.
There's a fourth insight that I think is underappreciated: the importance of the paper backup. Antarctic supply officers keep paper manifests. Smart preppers keep printed inventory lists. When the power goes out, when the cloud service is down, when your phone is dead — you still need to know what you have and where it is.
The spreadsheet is great until the battery dies. Then the clipboard is king.
Where does this leave us? The methods for planning extreme-environment logistics haven't changed much in fifty years. The fundamental math — consumption rates, person-days, buffer percentages — is the same math that was used to provision the early Antarctic expeditions. What's changed is the tools. Better forecasting software, better cold chain technology, better communication. But the core challenge remains: you're making bets about an uncertain future with lives on the line.
The environments are changing. Climate change is making resupply windows less predictable. Sea ice patterns are shifting. Permafrost is thawing under Arctic bases. The assumptions that worked for decades are becoming less reliable. How do you plan a buffer when the baseline is moving?
That's the open question that keeps logistics planners up at night. The rise of autonomous resupply — drones, robotic convoys, 3D printing of spare parts — could transform these environments in the coming decades. If you can fly a drone to the South Pole in winter, you don't need to stock nine months of everything. If you can print a replacement valve on demand, you don't need a warehouse full of Class IX parts. But the fundamental challenge of forecasting human needs isn't going away. People are still people. We still eat, get sick, break things, and need toothpaste.
Next time you run out of something at home, ask yourself: what would a submarine supply officer do? The answer is probably: they wouldn't have run out, because they did the math six months ago.
They have a paper backup just in case.
Speaking of things that don't run out, it's time for Hilbert's daily fun fact.
Now: Hilbert's daily fun fact.
Hilbert: During the Cold War, the Soviet Union operated the world's largest over-the-horizon radar system in the Namib Desert — a massive antenna array spanning over two kilometers, designed to bounce shortwave radio signals off the ionosphere to detect missile launches from the Southern Hemisphere. When it was operational, local farmers discovered their barbed-wire fences were picking up the signal and audibly playing Soviet radio broadcasts through the metal.
Barbed wire as a speaker. That's a new one.
That's going to stick with me.
This has been My Weird Prompts. Thanks to our producer, Hilbert Flumingtop. If you enjoyed this episode, leave us a review wherever you listen — it genuinely helps. We'll be back next week with another prompt from the void.
Until then, count your toilet paper and know your buffer percentage.
