When it comes to supply chain management, anytime you can take discrete pieces of work and group them together intelligently, you have the opportunity to optimize. The more time you have, and the more discrete pieces of work you can group together, the larger the potential savings from optimization.
I recently wrote about waving in the warehouse. Wave management is based on intelligently grouping together a batch of warehouse activities for later release to the floor. I’m a proponent of waving, but does it always make sense? Are there some things to watch out for when you wave?
First of all, not all business models lend themselves to waving. Here’s what Eric Lamphier, a Senior Director of Product Management at Manhattan Associates (an ARC client), told me in a recent conversation:
“The facilities that are more likely to flow orders directly from the ERP to the warehouse floor [essentially skipping or highly automating the wave] are those that have fewer physical and labor constraints and those that are trying hard to meet more aggressive service level agreements. I’ve seen this in ‘same day ship guarantee’ business models, high tech VMI (Vendor Managed Inventory) facilities that replenish raw materials to assembly operations, and in spare parts replenishment facilities.”
I would add that cross dock and highly automated flow through facilities also don’t lend themselves particularly well to waving.
Secondly, I’ve been on warehouse tours where I’ve seen some workers killing time (smoking out on the dock, talking to friends) until the next wave is released. One answer to this situation is visibility. Managers on the floor can spot idle workers and redeploy them. Further, today’s warehouse management systems (WMS) provide much better real-time visibility than in the past. By having real-time visibility of workers, managers can move employees to another work zone or assign them specific tasks. However, Eric thinks there is a better solution:
“We [Manhattan Associates] strongly recommend a more hands-off task interleaving configuration. Task interleaving takes the pressure off managers by automatically assigning tasks to users based on the completion of their previous task. You can configure the system to assign non-wave generated work, such as inventory put-away or cycle counting, and it can assign work from an adjacent work zone if the employee is authorized to do the work. Furthermore, the automatic task priority bumping logic ensures that the most important tasks are placed at the top of the queue without manager intervention.”
Wave engines are complex. I saw a presentation on the configuration options for Manhattan’s wave engine at its annual user conference a few years ago and I was struck by its power. But I also realized that learning how to configure the solution to maximize its effectiveness would take some time. Manhattan has since released a visual modeling tool that customers can use to create business rules—i.e., the tool can be configured by operational personnel, not just the IT staff.
Despite this visual modeling tool, I still believe waving is something a user learns how to use better – parameterize better – with time. Like the card games Bridge and Hearts, you don’t become a master over night.