Introduction
In my previous posts in this series on warehouse optimization I proposed methods for the assignment of items to storage locations also called the SLAP (see: MILP for SLAP and Genetic Algorithm for SLAP)
We solved the SLAP by considering the distance of the items to the depot and the affinity between two products. But how can we assess how good the solution of our algorithm is?
While we used an objective function that considers the distance to the depot and the affinity between two items, this value isn’t realy interpretable in an operational setting. In a real life warehouse setting we also have to consider the picking route as the time spend picking is of the greatest concern for a warehouse manager. Similar to the SLAP the so called Order Picking Problem is an NP-Hard problem and exact solutions are difficult to obtain. As we need fast solutions for large scale warehouses we again have to use a heuristic approach.
In this part we will develop a python based simulation and visualization tool to compare different routing strategies.
Roadmap
As the development of a simulation tool is a big task, we will split this in multiple parts. In the first part we will understand different routing heuristics and compare them. The second part will be concerned with building a data structure for our warehouse. In the third part we will construct the algorithm for our chosen routing heuristic and construct the simulation loop. In the final part you will learn how to leverage datastructures and an open source python visualization library to animate picking paths.
Routing Heuristics
Let us start with the comparison of two routing strategies that are wildly used in real life warehouse operations: S-shaped and Return Routing.
S-shaped routing
The S-shaped strategy, also referred to as the Traversal strategy, entails following a route through the warehouse aisles, ensuring thorough traversal of each aisle to be visited while skipping those where no items need to be picked. This results in a pattern resembling the letter S as the picker moves through the aisles. Starting from the left side of the warehouse, the picker enters an aisle from one end and exits from the other after picking the necessary items. Upon completing the picking, the picker returns to the front end of the aisle. This straightforward strategy is commonly used due to its simplicity and ease of implementation.
Illustratively, in a warehouse layout featuring a central depot and devoid of cross aisles, the order picker initiates from the depot and proceeds to enter the aisle with items nearest, whether on the leftmost or rightmost side. Moving from one item to the next, the picker traverses the entire aisle before eventually returning to the depot. [1]
Return routing
The return routing strategy is one of the easiest strategies for order picking. In this strategy the aisles are always entered from the front and left on the same side after picking the items. Thus it is called “return” strategy. Aisles that do not contain items to be picked are skipped. This strategy is especially useful when we slot the items in such a way that fast movers and items that are frequently ordered together are stored nearby.
Pros and Cons of both Strategies
S-shaped Routing:
Pros:
- Efficiency: By ensuring thorough traversal of each aisle while skipping unnecessary ones, the S-shaped routing strategy minimizes unnecessary travel time for pickers, leading to improved efficiency.
- Simplicity: The straightforward nature of this strategy makes it easy to understand and implement, reducing training time for new employees.
- Optimization: By starting from one end of the aisle and exiting from the other after picking items, the strategy maximizes the utilization of space within the warehouse aisles.
Cons:
- Limited Flexibility: S-shaped routing may not be suitable for all warehouse layouts, particularly those with complex aisle configurations or cross-aisle paths.
- Potential for Congestion: In warehouses with high traffic or multiple pickers operating simultaneously, the sequential movement through aisles may lead to congestion or bottlenecks.
- Inefficient for Certain Layouts: In warehouses with irregular aisle layouts or scattered storage locations, the rigid path of the S-shaped routing strategy may result in suboptimal picking routes.
Return Routing:
Pros:
- Simplicity: The straightforward nature of the return routing strategy makes it easy to implement and follow for order pickers, reducing the likelihood of errors.
- Optimized Aisle Navigation: By always entering aisles from the front and leaving from the same side, this strategy minimizes unnecessary movements and streamlines the picking process.
- Efficient for Specific Slotting: Return routing works well when items are slotted strategically, particularly for fast-moving items or those frequently ordered together, as it minimizes travel time between storage locations.
Cons:
- Limited Applicability: This strategy may not be suitable for all warehouse layouts, especially those with irregular aisle configurations or complex storage arrangements.
- Potential for Inefficiency: In warehouses where items are not slotted optimally, return routing may lead to inefficient picking routes, as pickers may need to backtrack or traverse long distances between storage locations.
- Dependency on Slotting: The effectiveness of return routing is contingent upon the organization of storage locations within the warehouse, requiring careful planning and management of inventory placement.
Selection of a Strategy
Given our current objective of constructing a simulation, it’s essential to narrow our focus to a single strategy. Considering our prior emphasis on enhancing storage location assignments in previous articles, the return routing strategy emerges as a fitting choice.
Conclusion
I hope you enjoyed this short comparison of two routing heuristics for order picking. The next part will be essential as we prepare the datastructures for our return routing simulation. Stay tuned!
