SCOR DS · Orchestrate · caselet
One disruption, five agents, one notebook: running the port-strike playbook
A port strike touches ordering, sourcing, and fulfillment at once. Orchestration is what keeps five agents solving one problem instead of five — decisions committed in 3 days, expedite spend capped at $120k.
The situation. A port strike stranding three inbound vessels doesn’t stay in one lane. It is simultaneously an allocation problem (who gets short-shipped), a sourcing problem (which inbound reroutes are worth paying for), and a fulfillment problem (which promises to re-cut). In 2021, the last event like this took the company three weeks of standing meetings to work through — decisions made serially, each one stale by the time the next was taken.
The signal. The risk feed raises the strike event; the Judgment Layer opens one disruption subject and pins it. Order Planning, Supplier Performance, Logistics, and Demand Planning agents each read the same fixed subject — same vessels, same SKUs, same customers — instead of four private versions of the truth. Four solvers and the Judgment Layer holding the envelope: five agents on one problem.
The decision. Orchestration is sequencing, not heroics: allocate against confirmed supply first (Order), then decide which reroutes change that supply picture enough to pay for (Source), then re-cut promises against the resulting dates (Fulfill) — with an expedite budget cap of $120,000 set once, at the top, instead of approved piecemeal.
The write-back. Each agent commits through its own systems — ERP allocations, rerouted POs, TMS promise dates — and every decision archives to the same notebook with its rationale and the telemetry it read.
The outcome.
| Metric | This event | 2021 baseline event |
|---|---|---|
| Signal → decisions committed | 3 days | ~3 weeks |
| Expedite spend | $120k (capped) | ~$310k (uncoordinated) |
| Decisions traceable to rationale | All | Meeting minutes |
The takeaway. SCOR DS makes Orchestrate a process of its own for a reason: in a real disruption, the expensive failure isn’t any single decision — it’s five good decisions made about five different versions of the problem.
Representative scenario; figures are synthetic. SCOR is ASCM’s framework. The allocation leg of this story is worked in detail in the ATP short-ship example.