VARGO® was recently featured in the July issue of Modern Materials Handling Magazine. Read the complete article.
The catch in going waveless
The “waveless” order fulfillment trend in warehouses is more than continuously releasing orders instead of batching work into waves—it’s an overall approach to metering and pacing the flow of work to optimize labor and equipment.
By Roberto Michel, Editor at Large
July 01, 2016
When you start looking into “waveless” order fulfillment many tech-laden terms crop up, like optimization, algorithms and warehouse execution system (WES) solutions. But peel away the technology, and waveless is really about a smoother, pull-based way of processing work in warehouses and distribution centers, say those involved with the trend.
“Waveless is a different way of thinking about things,” says Jay Moris, president of Invata, which offers WES software. “It’s letting certain resources pull available work into them, rather than having some predetermined master system that divides work into big chunks to try to push work through the system.”
The increased attention to waveless stems from the need to be more flexible in meeting today’s multi-channel order requirements while still being efficient in the use of automated materials handling systems and labor, according to Jim Barnes, president and CEO of consulting firm enVista. This flexibility need ties back to the fact that with traditional wave releasing under warehouse management system (WMS) solutions, waves are static batches of linked tasks that must be processed from start to finish before being able to accommodate new requirements. “
When a wave is released in the typical WMS, what the system is doing is creating a bunch of tasks, and those tasks are rigid,” says Barnes. “You are typically committed to finishing that batch of work before accommodating new orders. In a world where people are trying to do things in a more real-time fashion and compress order cycle times, that batch thinking becomes a constraint. So I think people in DC operations are starting to get attuned to the idea that large waves put constraints on the ability of a facility to react to demand signal changes.”
In response, says Barnes, we’re seeing interest in a more flexible, pull-based approach that borrows from manufacturing concepts like one-piece flow. But waveless or “continuous flow” methods are challenging to achieve in many DCs, adds Barnes, because of order variety and the use of sophisticated automated equipment such as large unit sorters for which work needs to be optimized at perhaps dozens of chutes.
So while there is interest in waveless, says Barnes, doing it correctly involves the use of WES functionality capable of doing things such as continuously releasing work to the floor and managing work assignments in a level/pull fashion. WES also typically encompasses warehouse control system (WCS) functionality—the layer of software that communicates with automated materials handling equipment. It also spans WMS-level functionality to sequence inventory replenishment from reserve locations to the areas that feed key pieces of automation. This makes waveless an issue that spans multiple software domains and tends to have the highest payoff for automated facilities, adds Barnes.
“When people get into the conversation between wave and waveless, I think there needs to be the distinction that waveless is really best suited to more of an automated environment,” says Barnes.
Method, not moniker
Art Eldred, client executive for systems engineering at VARGO®, a WES provider, says that to better understand the wave vs. waveless issue, it is useful to consider the nature of what constitutes a batch of work. In traditional wave releasing, work for multiple orders is batched together for efficiency in travel and other tasks, but batches are inflexible. If the batch calls for two items to be picked at a location but only one is available, batch logic bogs down the process because it wants to complete all picks before proceeding.
“In a waveless world, if you have two orders to pick at a location, but there is only inventory to fill one order, you would still pick that one item because you have enough inventory to complete that order,” says Eldred. “Thus, no single order is dependent on the other to complete a batch. As a result, you have eliminated non-essential relationships.”
Another key concept, adds Eldred, is that a WES supporting continuous flow may still make use of a batch of tasks to gain efficiencies, but it is a flexible, dynamic batch in which the intelligence of the WES constantly adds orders to the batch as downstream work is completed. “We think of a batch as a revolving batch of work, meaning that as one order is completed, another order goes back into the process,” says Eldred.
Vargo’s WES has been used for waveless operations by major retailers such as American Eagle and Forever 21. Vargo uses the term “continuous flow system” to describe these types of implementations.
Some warehouses may still be doing retail store fulfillment in full case or pallet loads where the more traditional wave approach works fine, says Eldred, but because of order fulfillment complexities tied to e-commerce and piece-based inventory management, continuous flow is gaining interest.
“Your work batches get much smaller because you have many different work tasks involved in picking and shipping out small e-commerce orders that are just more labor intensive and difficult to manage compared to filling a trailer with large retail orders,” says Eldred. “This drives a necessity for increased staging, gating and metering of work. A WES has the intelligence to manage these small work tasks to maintain efficiency throughout your omni-channel operations.”
The waveless term caught on a simple way to describe continuous flow for DC professionals who had become used to processing work in large waves, according to Marc Austin, group vice president of sales for Fortna, a consulting firm which offers WCS/WES software.
“When the first discussions started about continuous processing, it was confusing to people who said, ‘well, we work in waves,’ so the term ‘waveless’ was used to explain some differences,” says Austin. “But waveless is actually a small piece of what I would call overall system optimization, which is about getting the most out of what you have today or what you would like to build.”
The key with WES for continuous flow, says Austin, is that it has the WCS-level visibility and WES-level optimization logic to release work in small increments based on the availability of equipment assets and labor resources. So think of WES as a brain for pull-based, continuous processing of work, says Austin, rather than a point solution for waveless.
It’s about creating a continuous flow of orders, from start to finish, with the view of availability of downstream assets and labor resources, and taking into account dock-out and other rules, to create an efficient flow of work,” says Austin. “So it’s about optimizing the whole operation so that resources can call upstream and say, “I have availability,’ rather than trying to cope with peaks and valleys. So whether you want to work in a wave, or in a waveless environment, either one of those environments can be optimized, based on your systems, your assets and your processes.”
Many DCs continue to process orders as waves, especially for traditional channels with less each picking. However, because of omni-channel pressures and the need to accommodate tighter service levels, wave releasing is changing. Some providers say smaller waves in sizes matched to the work zone capacity and synchronized to the equipment and material flow can help.
“The idea is to balance the order fulfillment processes, so our WCS software continually evaluates work completion at pick zones and releases batches of orders matched to the zone work capacity,” says Dan Hanrahan, CEO of Numina Group, an integrator which offers a WES/WCS solution. “Essentially, we can release groups of orders based on the min/max order capacity across all the work zones as opposed to a strictly continuous flow release, which could over saturate downstream conveyor equipment.”
For Hanrahan, WES/WCS-based order releasing goals, regardless of batch size, should aim at a balanced flow of work around actual conditions on the DC floor. “What we are trying to do is keep the workers working at a steady pace matched to the zone work standards, measure the actual performance and ensure the materials handling automation system performs to the design rate without over assigning too much work to any one zone,” he says.
WES solutions typically have some type of order management function that is constantly making decisions about what work to release from an order pool. While WMS software may use rules to build waves, with WES, the solutions typically have WCS-based visibility that WMS systems lack, says Nancy Malone, a software and consulting manager with viastore, whose software includes WES.
“WCS/WES solutions are more dynamic in nature because they can provide sub-second decisions for automated routing and materials handling equipment,” Malone says. “Thus, offering this waveless order processing capability comes naturally to a well-engineered [WES] solution.” Invata’s Moris agrees that real-time visibility, combined with order release logic, is central to being able to properly release the right amount of work to the right areas. “We are only letting out the inventory and orders we need at any moment in time,” says Moris. “We are letting orders flow through the system in a way that is complementary to their ability to be quickly processed. There are no large queues.”
Companies such as Destination Maternity (see p. 16) have used Invata’s WES-like warehouse software to support balanced, continuous flow operations, says Moris. As part of this, a key WES function is the ability to reassign labor to areas such as busy sorter chutes. “WES needs to support dynamic and flexible personnel assignment,” says Moris.
To ensure material availability and avoid bottlenecks as part of waveless, a WES solution also should have replenishment management functionality, Moris contends. While a WMS might handle receiving or put away to bulk storage, a common demarcation line between WMS and WES functionality is replenishment to forward areas, with the WES governing replenishment. “When we are in control of inventory from door to door, then we can be much more capable,” says Moris.
WES provider Dematic Reddwerks defines waveless as a single-batch operation that continuously releases new orders from a constantly morphing pool of orders to downstream resources targeted to handle the volume. To support this, says Alex Ramirez, vice president with Dematic Reddwerks, WES functions should span order pool management, including rules around service levels; the ability to pace and manage workflows; the ability to manage both people and equipment; and inventory and replenishment management.
“With waveless, the system is looking at what the selectors are, what the order pool is doing, and weaving in work in real time to continuously flow product through the DC,” says Ramirez. “The goals are about equipment utilization, labor productivity, order accuracy, and ultimately, compressing the order cycle time.”
EnVista’s Barnes warns that users pondering WES to support continuous flow should look carefully at a WES vendor’s WMS-level capabilities around inventory replenishment because waveless typically involves complex sequencing of inventory. “You have to feed the beast through proper replenishment,” says Barnes. “Regardless of the solution architecture, you always have to ask, who is going to be the inventory system of record, and then start determining the solution architecture for waveless from there.”
For certain, concludes Barnes, the interest in continuous flow is real and has been building for several years as omnichannel complexities have become more pronounced and the constraints of rigid waves more apparent. “I think people are starting to think: Is there a better way?” says Barnes. “How can I optimize the use of my equipment, my labor, in more of a level pull fashion, so that as I have change in my demand signal, I can react to it?”