How AI Can Improve Picking

Picking is an essential part of daily warehouse activities. Because of its importance, organizing it efficiently directly affects the bottom-line. There are different "picking methodologies." In this Monday Learn Day post, I'll focus on "batch picking" or "multi-order picking."

What is it? In contrast to single order picking, in which pickers fulfil one order at a time, batch picking relates to the activity of tackling multiple orders at the same time. Efficiency gains are realized through "grouping" items on the picking list so as to create an optimal picking path. The main "win" for pickers is that they don't need to revisit the same picking location over and over again.

Crucial component of "batch picking " is a concept called "the batch factor." The batch factor refers to the method of consolidating multiple orders or tasks into a single "batch" for processing within the warehouse. Specifically, it measures how many orders or items are grouped together during picking, packing, or shipping. By grouping tasks, the batch factor seeks to minimize travel time, reduce redundancy, and maximize the throughput of a warehouse.

How Does Batch Picking Impact Warehouse Efficiency?

There are three main "wins" associated with batch picking. These are:

• Reduced Travel Time: In large warehouses, travel time often consumes a significant portion of the time spent by workers. By increasing the batch factor, workers can pick items for multiple orders in one trip, reducing the need for repeated trips to the same location. This can lead to a drastic reduction in total travel time, making operations more time-efficient.
• Better Resource Utilization: A higher batch factor optimizes the use of both labor and equipment. For instance, instead of a picker traveling with a half-full cart or forklift, they can move a fully loaded one, maximizing the capacity of every trip. Similarly, conveyor belts and sorting systems can be used more effectively when more items are processed at once, leading to better equipment utilization.
• Improved Order Throughput: Grouping multiple orders under a higher batch factor can lead to quicker order processing times. If workers handle multiple orders in one go, the time it takes to pick, pack, and prepare shipments is shortened, leading to faster fulfillment rates and potentially higher customer satisfaction.

How AI is Changing the Game

Optimizing batch picking in continuously changing warehouse operations can be tough. Due to its capacity to handle a large number of (changing) parameters, AI support agents can design optimal picking schedules. Particularly interesting in my view is the growing maturity of reinforcement learning applications. This subfield within the broader AI discipline excels at dealing with changing warehouse lay-outs and high order variability.

In case you'd like to do some more reading, the following article (open) by Cals et al. (2021) may be useful: https://www.sciencedirect.com/science/article/pii/S036083522100125X