Student project: Research on the efficient and dynamic scheduling of the incoming transport at a warehouse

The purpose of my research at the warehouse of Etos (Etos is part of Ahold) was to deliver a dynamic inbound transportation schedule which is flexible for tactical adjustments in the future.

The goal was to lower cost (both inventory and staff) and to create a more efficient receiving process (more deliveries with the same capacity). I have looked into past researches about inventory systems and ordering strategies, but the situation at this Dutch retailer was different. In the calculation of a right order strategy for an article, usually transportation costs are added to the total costs of a delivery. In this case transportation costs are included in the purchasing price. A higher delivery frequency could result in an increase in this purchasing price, but this differs per supplier (in total the retailer has over 300 suppliers). Therefore I have ignored this in the model. The recalculation of the order frequency resulted in a delivery frequency per supplier. In a second and third model I scheduled the suppliers on a certain day and at a certain time. The simulation model was used after these schedules were sorted.



The simulation model provided insight into the occupation rate and waiting times in different scenarios. The scenarios are different formats of incoming transportation schedules to test the effect of this format on the required capacity and reliability. The purpose of the model was to deliver a number of rules of thumb for the Dutch retailer. In the future, these basic rules can give guidance whenever strategic choices about the inbound transportation schedule have to be made. The rules of thumb are the result of the effect that the adjustment will have on the occupation rate and / or waiting periods, based on the simulation results. I will highlight some of these results:

1. The effect of reducing the number of hours in a working day from 14 to 7 hours: the average content of the queue hardly decreases and the average content of deliveries at the docks remains the same. The advantage of clustering the work in 7 hours is a decline of 1 hour of total working capacity per day. This results in a cut of 100 euros per week. The negative effect is that the maximum capacity of the required space rises from 4 to 7. The space is restricted and therefore this is not expedient.
   Rule of thumb: there is no advantage in decreasing the number of hours a day for this warehouse.
2. The impact of suppliers meeting their time of arrival: this causes a 60% decline in waiting time.
   Rule of thumb: Etos must ensure on time arrivals.
3. The effect of the proper organization of the discharge of pallets from the docks: when docks are held occupied for 10 extra minutes due to a bad organization of the discharge, the waiting time increases with over 700%.
   Rule of thumb: the discharge of pallets from the docks should start as soon as possible.
4. The effect of spreading the capacity evenly over the day: with an arrival list in ED I changed the number of available docks every hour. In the current situation the capacity regularly changes with 1 to 6 docks per hour. There is a wide variety in available capacity per hour. In the simulation this resulted in overcapacity (occupation rates > 100%) and long waiting times. I built a model for the warehouse that spreads the incoming trucks evenly over the day with a maximum variety of 1 dock between the hours.
   Rule of thumb: Etos should try to spread the arrivals evenly over the day using the model.

Anne Goudsmit MSc in Operations Research University of Amsterdam.