Supply-chain disruption planner

Ingest disruption signals (port closures, supplier delays, weather, strikes), match them to affected purchase orders and shipments, then run parallel branches: find alternative suppliers, alternative routes and inventory reallocation. A simulation node computes cost, delay and service impact with code. The planner ranks options, explains trade-offs and pauses for a planner's approval before updating orders and notifying customers.

Join the disruption's location, supplier or lane with open purchase orders, shipments in transit and bills of materials, then propagate to finished goods and customer orders that depend on them, prioritised by due date and value.

Score each option by total cost (freight, expediting, penalties), delay against customer due dates, service level, risk and capacity, with weights agreed by the business, and show the top options with their trade-offs rather than one answer.

Numbers must be correct and reproducible. Code applies real rates, transit times, inventory levels and constraints the same way every time. The model is good at understanding the situation and proposing options; the calculation belongs in deterministic tools.

Costs, lead times and inventory projections must be exact and reproducible; code uses real rates and constraints, while the model explains results and drafts communications.

Use contracted rate tables, carrier APIs and historical lane performance, with freshness dates, rather than letting the model estimate prices.

Subscribe to carrier and port status feeds, supplier updates, news and weather APIs, classify relevance with filters and a model, deduplicate, and score severity before triggering planning.

Check available stock, safety stock and transfer lead times, simulate the effect on each region's service level, and propose transfers that protect priority customers.

As soon as a delay to their committed date is confirmed (after planner approval), with the new estimated date and options, rather than waiting until the original date passes.

On-time delivery under disruptions, cost of recovery versus baseline, time from signal to plan, planner acceptance rate and customer complaints.

Model scenarios (short, expected, long), compute options for each, and prefer robust choices or staged decisions that can be revised as information arrives.

A detection node turns signals into a disruption record, an impact node finds affected parts, orders and customers through the bill of materials, parallel branches generate options (expedite, reroute, reallocate, alternate supplier), a simulation node costs each with code, then an approval interrupt waits for the planner before writing changes back.

Score each option on total cost, number of late orders weighted by customer priority and penalties, and risk, show the trade-offs side by side, and let planners set weights per product line; never pick automatically when the options trade revenue against cost.

Score each on total cost, delivery delay, stock-out risk and customer priority, with weights the business sets. Show the trade-offs side by side, highlight the recommended option and explain why, and let the planner choose.

Reliable and early: carrier status feeds, port and weather alerts, supplier notifications and news, each with a location and expected duration that can be matched to routes and orders.

Read purchase orders, inventory and BOMs from the ERP (SAP, Oracle, Dynamics) via APIs, and write approved changes back through controlled transactions with audit trails.

Search approved supplier lists by part or material, check qualification status, capacity, price and lead time, and flag suppliers that need qualification before use.

Affected orders, ranked options with cost and delay, assumptions, risks, and the customer communications that will be sent, with buttons to approve, modify or request more options.

Set severity thresholds by impact on orders and customers, batch low-impact signals into daily summaries, and only trigger full planning for material risks.

Replay past disruptions with historical data, compare recommended options with what planners chose and the actual outcomes, and test edge cases like multiple simultaneous disruptions.

Aggregate affected orders across disruptions, solve allocation jointly where resources are shared (inventory, capacity), and prioritise by customer and margin.

Extra inventory held to absorb variability in demand and supply; the planner uses it as a buffer but should flag when using it risks future shortages.

Show the calculation inputs, assumptions and the reasons an option ranks higher, in business language with links to data, so planners can challenge them.

Record the received quantity, recompute shortages for the missing part, allocate the received stock to the highest-priority orders first, and trigger options only for the remaining gap.

Estimate emissions per option from mode, distance and weight using standard factors, show them next to cost and service impact, and let policy cap high-emission options such as air freight unless the business impact justifies it.

Combine the baseline forecast with known effects such as customers pulling orders forward or substituting products, use signals from order intake and sales teams, and show scenarios rather than a single number when uncertainty is high.

Check every option against constraints in code: capacity at plants and warehouses, carrier availability, supplier qualification, minimum order quantities and customs lead times, and drop options that violate them before they reach planners.

The list of components and quantities needed to make a product; it is how a part shortage is traced to affected finished goods.

Sync ERP and logistics data frequently, show data timestamps in plans, and re-run plans when key inputs change.

Start with one region or product line in advisory mode, compare with planners' decisions, then expand and allow more automated steps.

Signals received, affected orders identified, options computed, the chosen plan, approvals and changes written to the ERP.

Publish agreed changes through the existing portals or EDI messages, send revised dates and quantities, and record acknowledgements, so partners work from the same plan.

1. AGuesswork
2. BContracted rate tables and carrier APIsCorrect answer
3. CThe model's estimate
4. DSocial media

Why: Prices must be exact.

1. AThe customer's consent
2. BA human planner's approvalCorrect answer
3. CThe model's confidence
4. DThe supplier's signature

Why: Approval before changing orders.

1. AAfter the original date passes
2. BNever
3. COnly if they ask
4. DAs soon as a delay is confirmedCorrect answer

Why: Early notice builds trust.

1. AThe model's memory
2. BLive carrier quote APIsCorrect answer
3. CLast year's slides
4. DCustomer emails

Why: Rates change constantly, so they must come from live sources.

1. AResults must be exact and reproducibleCorrect answer
2. BModels cannot read numbers
3. CCode is more creative
4. DIt is cheaper always

Why: The model explains results.

1. AOne answer only
2. BRandomly
3. CWithout costs
4. DTop options with trade-offsCorrect answer

Why: Planners choose with context.

1. AThe invoice
2. BThe org chart
3. CThe website
4. DThe bill of materialsCorrect answer

Why: Propagate impact through BOMs.

1. AIgnore it
2. BWait until it ends
3. CScenarios with robust or staged choicesCorrect answer
4. DAssume the best case

Why: Revise as information arrives.

1. ANumber of signals
2. BModel size
3. CReport length
4. DOn-time delivery under disruptionsCorrect answer

Why: Track it alongside recovery cost.

1. AIgnored entirely
2. BSent to customers
3. CBatched into summaries below a thresholdCorrect answer
4. DFully planned every time

Why: Thresholds avoid alert fatigue.

1. AThe model's estimate
2. BA fixed percentage
3. CSupplier promises only
4. DA code-based simulationCorrect answer

Why: Costs must be reproducible and correct.

1. AMade automatically by the cheapest rule
2. BRandom
3. CLeft to suppliers
4. DMade by the planner with trade-offs shownCorrect answer

Why: Business priorities need human judgement.

1. AThe carrier
2. BThe customer
3. CThe impact agent
4. DA human plannerCorrect answer

Why: The planner owns the decision and its cost.

1. AImmediately after the disruption
2. BAfter the plan is approvedCorrect answer
3. CNever
4. DBefore options are priced

Why: Sharing unconfirmed dates causes confusion and lost trust.

1. ALogo colour
2. BWebsite design
3. CNothing
4. DQualification, capacity and lead timeCorrect answer

Why: Unqualified suppliers add risk.

1. AThe ERP with audit trailsCorrect answer
2. BA spreadsheet only
3. CThe model
4. DEmail

Why: Controlled transactions.

1. ABy reading prompts
2. BNot at all
3. CReplaying past disruptionsCorrect answer
4. DOnly in live crises

Why: Compare with real outcomes.

1. AReturned items
2. BBuffer inventory for variabilityCorrect answer
3. CDamaged goods
4. DMarketing samples

Why: Using it may risk future shortages.

1. ANo planning
2. BRandom choice
3. CJoint allocation of shared resourcesCorrect answer
4. DSeparate plans ignoring each other

Why: Shared resources overlap between disruptions.

1. ANothing
2. BInputs, assumptions and reasonsCorrect answer
3. COnly the final choice
4. DModel weights

Why: Planners can challenge them.

1. AOnly email
2. BOnly phone calls
3. COnly social media
4. DCarrier, port, supplier, news and weather feedsCorrect answer

Why: Classify relevance and severity.

1. ATransfer lead times and regional service levelsCorrect answer
2. BOnly the cheapest warehouse
3. CNothing
4. DWarehouse names

Why: Protect priority customers.

1. AThe newest orders
2. BNobody until the rest arrives
3. CThe highest-priority orders firstCorrect answer
4. DOrders in alphabetical order

Why: Priority allocation limits the damage.

1. AThe supplier's name
2. BOrder value only
3. CThe planner's location
4. DMode, distance and weightCorrect answer

Why: Standard factors convert transport into emissions.

1. APlanners noticing later
2. BCustomers
3. CConstraint checks in code before reviewCorrect answer
4. DAsking the model politely

Why: Planners should only see workable plans.

1. AScenarios when uncertainty is highCorrect answer
2. BOne exact number
3. CLast year's figure
4. DZero

Why: Scenarios make uncertainty explicit.

1. AModel names only
2. BRandom IDs
3. CData timestampsCorrect answer
4. DNo dates

Why: Stale data misleads planners.

1. AOne region or product line in advisory modeCorrect answer
2. BEverywhere at once
3. CWith full automation
4. DNowhere

Why: Expand after comparison.

1. AOnly the final email
2. BNothing
3. COnly errors
4. DSignals, options, approvals and ERP changesCorrect answer

Why: This gives full traceability.

1. APhotoshop
2. BFigma
3. CSlack only
4. DSAP, Oracle or DynamicsCorrect answer

Why: Read orders and inventory.

1. AOnly one option
2. BUI themes
3. CBilling
4. DSuppliers, routes and reallocation togetherCorrect answer

Why: Faster option generation.

1. AExisting portals or EDI messagesCorrect answer
2. BPersonal phone calls only
3. CSocial media
4. DNothing

Why: Partners need the same plan in their systems.