Gains:
- Ability to combine demand, stock, warehouse, route and supplier modules in an end-to-end workflow and use artificial intelligence as a holistic decision support layer
- Ability to understand and draft elements of AI governance policy (data classification, human approval points, audit trail, ethical principles)
- While monitoring carbon footprint and sustainability indicators with the support of artificial intelligence, we can ensure that the responsibility and final decision remain with the human.
Module Exam
1. Which of the following is the most accurate positioning for artificial intelligence in logistics and supply chain?
- A) Artificial intelligence can finalize shipping orders and purchase orders without human approval
- B) Artificial intelligence only works in writing text, it has nothing to do with operational data
- C) Artificial intelligence is an assistant and decision support tool; Responsibility and final approval of critical operational decisions lie with humans ✔
- D) Artificial intelligence always makes more accurate decisions than humans, so auditing is unnecessary
Description: Artificial intelligence is an assistant and decision support tool that produces predictions, drafts and analysis. Responsibility and final approval of decisions affecting money and operational continuity, such as shipment, stock investment and supplier selection, belong to the competent expert and manager.
2. We know that a demand forecasting model is trained with data from the past 3 years. A brand new product will be launched on the market. What is the status of the AI forecast for the first month demand of this product?
- A) Forecast for new product with no historical data is poorly based; Similar product analogy and expert judgment take precedence ✔
- B) Estimation for new product always gives the most accurate result
- C) The model automatically learns the new product correctly, no data is needed
- D) It is impossible to forecast demand for new products, no method can be used
Description: AI prediction is based on past patterns. For a new product without historical data, the model lacks a reliable basis; Similar product analogy and expert judgment come to the fore, and the prediction is approached cautiously.
3. What is the main purpose of the concept of safety stock?
- A) Keeping the tank fully filled at all times
- B) To facilitate accounting records
- C) Extending the shelf life of products
- D) Creating a buffer against uncertainty in demand and supply time and preventing stock depletion ✔
Explanation: Safety stock is additional stock held as a buffer against uncertainty in demand and supply time. Its purpose is to maintain the targeted service level by preventing stockouts during fluctuations.
4. What do group 'A' products generally refer to in ABC analysis?
- A) The cheapest and most diverse products with low value
- B) Critical products that are few in number but make up the majority of the total value ✔
- C) Dead stocks that see no movement in the warehouse
- D) Only imported products
Description: ABC analysis classifies products by value. Group 'A' are critical products that are small in number but account for the majority of total value/turnover; They deserve the strictest monitoring and control (Pareto principle).
5. What should the operations manager check before implementing a warehouse picking route suggested by artificial intelligence?
- A) Color and font of the route
- B) Only check whether the route is the shortest
- C) Site safety, ergonomics and compliance with actual warehouse constraints ✔
- D) Nothing; The AI route is always correct
Description: Artificial intelligence can suggest a mathematical route that shortens the distance; However, field safety, aisle one-wayness, heavy product ergonomics and actual warehouse constraints may not be fully reflected in the model. That's why field authenticity and security verification is essential.
6. What does the 'time window' constraint mean in the vehicle routing problem (VRP)?
- A) Allowed time interval during which a point can receive service for delivery/reception ✔
- B) How long does it take for the vehicle's tank to be filled?
- C) Driver's total working years
- D) Time when the fuel price changes
Explanation: A time window is the allowed time period during which a delivery or pickup point can receive service (for example, the grocery store only accepts goods between 08:00 and 11:00). The route must comply with these windows; AI must build the scenario under this constraint.
7. In predictive maintenance, artificial intelligence indicates a high probability that a vehicle's bearing may fail soon. What is the right approach?
- A) Ignoring the warning altogether, because AI is fallible
- B) Scrapping the vehicle immediately and without question
- C) Dismissing the driver without warning
- D) The technical expert confirms the estimate with a physical examination and makes the maintenance decision ✔
Description: Predictive maintenance is an estimate of possibility. Instead of immediately scrapping the vehicle or ignoring the warning, the technical expert confirms the prediction with a physical inspection and makes the maintenance decision. AI helps prioritize, not make the decision.
8. Why should 'single sourcing' be carefully monitored in supplier risk management?
- A) Working with a single supplier is always the cheapest, there is no risk
- B) Critical material is dependent on a single supplier, an interruption in that supplier can stop the entire chain ✔
- C) Single source is a problem because it only increases the number of invoices
- D) Single source risk is only valid for imports, not domestically.
Explanation: Single source risk is when a critical material is dependent on a single supplier. An interruption at that supplier (bankruptcy, natural disaster, strike) can stop the entire supply chain. Artificial intelligence can demonstrate this concentration; The alternative resource plan is the responsibility of the manager.
9. Artificial intelligence automatically extracted the amount, type of goods and GTIP code information from an import invoice. What should be done before transferring this data to the customs declaration?
- A) Transferring directly to the system, because artificial intelligence does not make mistakes
- B) Checking only the amount and skipping other fields
- C) The responsible expert compares the extracted data with the original document, verifies and approves it ✔
- D) Deleting the invoice and requesting it to be reissued
Explanation: Data extracted by OCR may be incorrect (wrong number, mixed code) in customs, tax and legal fields. It should not be processed without human verification; The responsible expert must compare it with the original document and confirm it.
10. In a control tower setup, artificial intelligence generates warnings for ETA (estimated arrival) deviations. On what does the reliability of these warnings depend most?
- A) Accuracy and up-to-dateness of the data it feeds ✔
- B) How many words are in the warning message?
- C) Depending on the display colors of the control tower
- D) Driver's phone brand
Description: The AI alert depends on the quality of the data it is fed. If the GPS signal, current traffic and accurate shipping status are missing or delayed, the summary and alert produced will be misleading. The principle of 'garbage in, garbage out' applies.
11. In transportation tender evaluation, artificial intelligence recommends the lowest bid. Which is correct when making the decision based on this suggestion?
- A) The lowest bid should be automatically selected in all circumstances
- B) No criteria other than price should be taken into account
- C) Artificial intelligence should finalize the decision on its own
- D) In addition to price, capacity, performance and continuity should be evaluated, competition rules should be followed, and the final approval should be with the manager ✔
Remark: The lowest price alone may not be the best choice; The carrier's capacity, delivery performance, insurance and continuity are also important. In addition, compliance with competition rules and final approval belong to the manager. Artificial intelligence speeds up the comparison, not the decision.
12. What does the MAPE (Mean Absolute Percent Error) indicator measure in a demand forecast?
- A) Total square meters of the warehouse
- B) On average, how many percent do the predictions deviate from the actual values?
- C) The volume of a vehicle's fuel tank
- D) Number of suppliers
Description: MAPE is an accuracy indicator that measures the average percentage deviation of predicted values from actual values. Lower MAPE means more accurate prediction; Knowing this deviation, the planner adjusts the safety stock and scenarios.
13. What is the best approach when giving operational and commercial data (price lists, supplier contracts, customer volumes) to an artificial intelligence tool?
- A) Freely use any tools, data sensitivity is irrelevant
- B) Pasting trade secrets into the most popular free tool
- C) Using institution-approved and confidentiality-guaranteed vehicles, masking/anonymizing sensitive data unless necessary ✔
- D) Transferring the data to the tool that offers the most features, without checking it at all
Disclosure: Business data is competitively sensitive and requires confidentiality. Institution-approved tools with data privacy guarantee should be used; Sensitive data should be anonymized or masked unless necessary; Trade secrets should not be included in vehicles that are open to the public and whose data goes to education.
14. Within the scope of sustainability, artificial intelligence suggests reducing carbon footprint by shortening routes. What is true about this suggestion?
- A) AI recommendation is based on assumptions; The final decision and responsibility lies with the human being, taking into account delivery, driver rights and customer restrictions ✔
- B) Artificial intelligence recommendation for carbon reduction should be implemented without question
- C) Sustainability is irrelevant to logistics and is not taken into account
- D) Route shortening has nothing to do with carbon
Description: Artificial intelligence can generate valuable recommendations for carbon and cost savings; however, the recommendation is based on assumptions and data. The final decision and responsibility remains with the human, taking into account constraints such as delivery commitments, driver rights and customer satisfaction.