The shipping schedule is highly uncertain, and the fuel supply plan changes frequently.
Complete English presentation of the source project analysis, preserving the full narrative, tables, figures and evidence boundaries.
Maritime Energy Customer Case · Pain Point 1
The shipping schedule is highly uncertain, and the fuel supply plan changes frequently.
From Static Nomination to ETA Probability and Event-Driven Scheduling
| North Star Metric: On-time and Complete Oil Supply Rate and Dispatch Cost per Ton |
|---|
Executive Summary
This case revolves around 'highly uncertain shipping schedules and frequently changing fuel supply plans.' It is based on a real maritime energy business chain and covers inquiries, nominations, scheduling, delivery, measurement, sampling, quality, settlement, claims, and continuous improvement. The goal is to transform the uncertainties at ports and on vessels into measurable and controllable service capabilities.
| Business Profile | Case-based parameters |
|---|---|
| Scene | The container ship supplied 1,200 tons of VLSFO and 150 tons of MGO at the hub port. |
| Time window | ETA changes continuously, berthing windows and barge resources are linked |
| Constraint | Oil inventory, barge tank capacity, tides, docks, customs, and safety windows |
| Loss | Empty sailings, waiting, port changes, panic buying, breach of contract, and vessel delays |
Failure signal
ETA enters the plan as a single point value, without P10/P50/P90
Information about shipowners, agents, terminals, suppliers, and barges is updated via email
Inventory and barge scheduling were not jointly optimized according to feasible windows
1. Representative Assignments/Transaction Examples
The following numbers are hypothetical professional cases used to illustrate economic and risk transmission, and do not represent the audit facts of any specific client.
Representative Economic/Operational Impact of This Pain Point
The source-document visual is presented here as an English-native analytical frame. The adjacent English narrative and tables preserve the full evidence and quantitative context.
| Project | Value/Impact | Explanation |
|---|---|---|
| Planned Contribution | +18 | ten thousand US dollars |
| ETA delayed by 36 hours | −4.2 | Barge Waiting/Rearrangement |
| Modify port supply | −3.8 | Cross-harbor transfer |
| Temporary procurement | −2.6 | Spot premium |
| Customer Delay Compensation | −1.5 | Service failed |
| Final Contribution | +5.9 | Profit shrank by 67% |
2. Root Cause Analysis
ETA enters the plan as a single point value, without P10/P50/P90
Information about shipowners, agents, terminals, suppliers, and barges is updated via email
Inventory and barge scheduling were not jointly optimized according to feasible windows
The change did not trigger procurement, pricing, credit, or customer notifications
Control Failure Chain
| Level | typical gap | Consequence |
|---|---|---|
| Business commitment | The quote/SLA does not reflect on-site constraints | Profits and services are inherently overestimated |
| Job execution | The event did not trigger reorganization, work stoppage, or escalation | Abnormal accumulation |
| Data evidence | Object, time, version, and signature are inconsistent | Unable to provide evidence |
| Organizational Governance | Cross-party responsibilities and the right to stop are unclear | Repeated failure |
3. Solution: Five-layer closed loop
| Hierarchy | Core Competence |
|---|---|
| 1 Voyage Event Layer | Integration of AIS, port queues, agents, weather, and terminal status |
| 2 Probability ETA | Output berth/completion time distribution and confidence |
| 3 Constraint Scheduling | Joint optimization of oil products, barges, ports, operation teams, and windows |
| 4 Dynamic Commitment | Show the client the commitable range and change costs |
| 5 Exceptional Scripts | Delay, port change, cancellation, and emergency refueling automatically trigger actions |
End-to-end operating mechanism
Ship/port/order events enter the unified voyage data mainline.
Feasibility, risks, economic impact, and evidence requirements of rule and model calculation services.
The person in charge shall carry out, stop, or escalate according to the authorization, and synchronize with the customer.
Delivery results, samples, measurements, costs, and claims are fed back to form a learning loop.
4. Implementation Roadmap
| Stage | Time | Delivery | Acceptance |
|---|---|---|---|
| Standardization | January–March | Caliber, SOP, Master Data, Responsibility Matrix | 20 assignments can be fully replayed |
| Visualization | March–June | Voyage/Order Control Tower, Exceptions and Evidence | Critical state T 0 |
| Intelligent | June–December | Prediction, Optimization, Risk Scoring, and Recommendations | Core KPI has entered the target range |
| Scaling | 12–18 months | Multi-port replication, supplier/customer collaboration | Standard coverage ≥90% |
The first 90 days
Select two core ports and one high-frequency fuel/service scenario.
Replay 20–30 normal, abnormal, and claim operations and establish a baseline.
First, verify the controllable loop manually, then gradually automate it.
Review false positives, missed alerts, adoption rate, economic value, and security impact every two weeks.
5. Outcome Indicators and Value
| Indicator | Baseline | 12-month goal |
|---|---|---|
| ETA MAE | 18 hours | ≤6 hours |
| On-time fuel supply rate | 78% | ≥95% |
| Barge running empty/waiting | Baseline 100 | Down 30% |
| Change Response | 4 hours | ≤20 minutes |
| Proportion of emergency procurement | 14% | ≤5% |
| North Star Metric: On-time and Complete Oil Supply Rate and Dispatch Cost per Ton |
|---|
Value structure
Operational value: on time, full quantity, qualified, complete evidence.
Economic value: Reducing waiting, claims, rework, and leakage of funds and services.
Risk Value: Reducing high-consequence risks to safety, quality, compliance, and asset bets.
Customer Value: Provide consistent services that are commitable, explainable, and auditable.
6. AI Evolution and Control Boundaries
AI Applicability
Predicted ETA and docking completion time
Rolling Optimization of Barges/Inventory/Team Scheduling
Assess the marginal cost of port changes and delays
Generate draft of customer notification and alternative plan
Control boundary
AI is responsible for extraction, correlation, prediction, and plan comparison, and does not replace the captain, on-site person in charge, inspector, or compliance officer.
All recommendations should display the data source, time, assumptions, confidence level, and failure conditions.
Safety/environmental red lines, legal requirements, and the finality rules of contracts must not be modified by the model on its own.
Automatically downgrade to manual operation in case of data interruption, significant changes, or model conflicts; retain the emergency stop authority.
7. Professional Conclusions and Next Steps
| Final judgment: Shift from static nomination to ETA probabilities and event-driven scheduling. Only when the 'on-time and complete oil supply rate and scheduling cost per ton' stably enter the target range, and no short-term profit is exchanged at the expense of safety, quality, compliance, or future options, is the capability truly replicable. |
|---|
Recommended next step
Conduct a 6-week diagnosis to create a list of events, losses, evidence, and control gaps.
Select a single-port/double-port pilot for 90 days to validate processes, data, and KPIs.
Include the final economic contribution together with safety, quality, and customer outcomes in performance.
After verification, copy according to port capacity tiers, rather than a simple 'one-size-fits-all' approach.
Applicable Caliber and Limitations
This report uses anonymized, case-based data. ISO 8217, MARPOL Annex VI, measurement standards, EU ETS, FuelEU Maritime, and port safety and environmental protection requirements should be based on the versions in effect for the applicable year and jurisdiction as agreed in the contract, and confirmed by qualified maritime, inspection, legal, compliance, or carbon professionals. This report does not constitute legal, classification, inspection, or regulatory advice.