INDEPENDENT POLICY & RESEARCHConceptual framework Edition 0.16.4
Resilience Lab · Fictional exercise

Trade and fuel continuity exercise

A civilian supply coordinator needs to know whether cargo can reach a receiving point and whether local diesel stocks can cover dispatch until the next receipt. These are two separate examples using fictional quantities.

Can two routes meet the receiving point’s demand?

Two routes can carry 60 and 50 crates per day. Both feed the same receiving point, which can handle 80 crates per day. Daily demand is 100 crates.

Possible delivery = min(60 + 50, 80, 100) = 80 crates per day.

What the result means. The receiving limit leaves 20 crates per day above possible delivery. The coordinator can investigate receiving capacity before expecting another upstream route to help. The result is a throughput ceiling. It does not establish that any cargo has departed or arrived.

What happens if the second diesel receipt is late?

The depot starts with 100 barrels of usable finished diesel. The first day brings 20 barrels. Demand is 90 barrels and dispatch capacity is 70. The exercise carries 20 barrels of unmet demand into day two. On day two, new demand is 40 barrels and the depot can dispatch up to 100. A separate, accepted receipt of 10 barrels is due. Each day’s receipt is available before dispatch when it arrives on time.

Fictional two-day diesel dispatch ledger, in barrels
Case and dayOpening stockReceivedNew demandBacklog at startDispatchedBacklog at closeClosing stock
On-time receipt, day 110020900702050
On-time receipt, day 2501040206000
Late receipt, day 2500402050100

Day two on time = min(50 + 10, 40 + 20, 100) = 60 barrels dispatched. Day two late = min(50 + 0, 40 + 20, 100) = 50 barrels dispatched.

What the result means. The late case closes with 10 barrels of unmet demand. Both cases end with zero stock, so ending stock alone hides the difference. An accepted offer stays out of usable inventory until it is physically received. The 10-barrel change is a receipt-timing test and does not measure a cooperation or AI effect.

To reproduce the late case, change day two’s received value to zero. Leave its offered and accepted amounts at 10.

What changes under the three arrangements?

Hold routes, receiving capacity, demand, staffing, budget and information release time equal. The three options use different procedures for handling the same records.

Equal-input comparison of civilian arrangements
MeasureFragmented arrangementsStructured non-AI cooperationMAR: AI-assisted cooperation
How teams handle informationOperators reconcile each other’s updates separately when discrepancies are found.Operators use a common worksheet and designated human reviewers to reconcile route and receipt records.AI can flag mismatches in approved records. Human reviewers confirm the evidence and each operator retains control.
Possible delivery at stated demand80 crates per day80 crates per day80 crates per day
Demand above possible delivery20 crates per day20 crates per day20 crates per day

To test whether the procedure matters, run the same timed exercise under each arrangement. Record duplicated offers caught, verified receipts, correction time, staff effort and false alerts. Report those measures beside service continuity. Require lawful authorization and product verification before relying on a proposed real shipment.

Which evidence would change the conclusion?

Check the receiving limit against observed unloading and release rates. Compare dispatch totals against operator logs and actual usable receipts. Review whether unmet requests can truly carry forward and whether stock quality, storage limits or expiry matter. A result based on finished diesel cannot be inferred from crude oil offers.

The May 2021 Colonial Pipeline case provides historical context on alternative supply routes and practical limits. These two fictional ledgers are independent of that case. Read the sourced fuel distribution case