Warehouse is an essential part of any distribution network of any goods from fruit to complex equipment. Supply chain and operations consulting experience shows relatively low maturity of Russian warehouse operators. The quality of warehousing service is often below expectations of both international and Russian clients. So, demand on improving warehouse operational efficiency is growing up.
Modern warehouse is a complex automated system functioning under permanent operational workload. This makes warehouse a very difficult object to manage, requiring thorough analysis to make any decision at design stage as well as during operation.
For example, a warehouse designer may be in a situation to make decisions like the following:
- choose between wide-aisle and narrow-aisle racks;
- for each storage zone make a choice between regular, push-back and gravitational racks;
- select appropriate warehouse equipment: conveyors, forklifts, etc.
During warehouse operation it can often be necessary to adjust the warehouse processes, for example:
- change a method of allocation pallets in storage zone;
- change the pickup and putaway orders dispatching algorithm.
Validation of such complex decisions is not an easy task as it requires consideration of many heterogeneous factors at a time.
Simulation is a method of complex systems behavior analysis based on using a computer representation of the system. Simulation model reproduces both structure and dynamics of the real object which allows to execute experiments considering virtually all significant aspects of the system. Simulation allows to find solutions by analysis of different cases (scenarios) and sometimes without a strict formalization of a problem.
All named features make simulation effective and efficient for our “warehouse” problems. Like no other method, simulation model can consider the following factors at once:
- forklifts routing and dispatching rules,
- putaway and pickup policies,
- warehouse layout details,
- technical characteristics of the equipment (moving speed, size, lifting speed, capacity, etc.)
Our experience tells that simulation is efficient and value-adding in the following cases:
- projects on warehouse design and operational model development;
- projects on warehouse operational performance improvement by means of redesigning processes and starting to use new equipment.
Let’s consider the example of using simulation to validate efficiency of using different putaway policy. So, some company operates a packed food retail warehouse. Deliveries are made by city’s retailers every morning from 6 to 10 am. The rest period of the day warehouses just receives shipments. Each client (retailer) strictly requires to make deliveries at a certain time in the morning. So, mornings are the hell for warehouse workers, they try to do their best but nevertheless orders are often delayed and only 85% of all deliveries leave the warehouse on time. Clients are very upset with the situation and starting to consider using another warehouse provider.
Assessment showed that goods are placed into the first available cell in the main storage zone. Meanwhile stored goods are very diversified by their turnover. Inability to extend the warehouse makes the situation even worse.
So what if we try to increase the throughput capacity of our warehouse by introducing a turnover-based putaway policy: the more likely the good is needed the closer it will be placed to assembly zone.
We’ll use a simulation model to analyze what we can achieve just put placing pallets in a different way. Our model should answer the following question: to what extent is it possible to decrease outbound order time by implementation of turnover-based putaway policy at warehouse?
Model will determine the following operation characteristics of the warehouse: average outbound order time, daily mileage of forklifts, warehouse zones utilization and fraction of shipments made within contracted outbound window.
Simulation (see the model below) showed that using turnover-based putaway policy allows to decrease average outbound order time by 6.4 % and thus to increase shipments within outbound window fraction from 87% to 94%. Transitional period lasted for 4 days and didn’t bring any problems with warehouse performance.
Demo model summary
|Name||Comparison of putaway policies at warehouse|
|Business problem statement||To what extent is it possible to decrease outbound order time by implementation of turnover-based putaway policy at warehouse?|
|Model input||Warehouse layout, outbound time window, rate of goods arrival, rate of outbound orders arrival, stock structure by turnover (ABC) categories, putaway policy (first available cell or turnover-based)|
|Model output||Average outbound order time, daily mileage of forklifts, warehouse zones utilization, fraction of shipments made within outbound window|
|Modeling method||Combined (agent-based and discrete-event) simulation|
|Modeling result||Using turnover-based putaway policy allowed to decrease average outbound order time by 6.4 % and thus increase shipments within outbound window fraction from 87% to 94%. Transitional period lasted for 4 days and didn’t bring any problems with warehouse performance|