The path to automated order picking at Cutter & Buck’s fulfillment center in Washington state began in its legacy pick module during the pandemic, when it was highly challenging for the apparel company to add peak season labor. First in 2020, and again in the holiday peak in late 2021, the combination of surging e-commerce sales and the labor availability constraint prompted Cutter & Buck leadership to pitch in directly to help.
Sales representatives, the CFO, and CEO Joel Peter Freet all rolled up their sleeves and helped fulfill orders in the three-level pick module. That lead to a realization that a more efficient order picking method was needed long term—one that could handle peaks with ease and that could keep its embroidery machines supplied with the right goods to customize on a close to just-in-time basis.
“We all pitched in, getting trained on our picking processes and committing to multiple shifts per week during the holidays,” Freet recalls. “By holiday 2021, our e-commerce was growing so fast we had no choice but to pitch in again. During our post-mortem, we explored ways to make the existing processes more efficient, including changing the layout of the picking module, or possibly reducing SKUs. In the end, we uncovered that to keep up with our growth, we would need a breakthrough in productivity to support our peak season sales in the future, and that’s what brought us to looking at next-generation picking automation systems.”
Cutter & Buck is an omnichannel retailer and supplier of active sportswear and lifestyle apparel. The Seattle-based company has partnerships or licenses for fan apparel with major professional sports leagues including the NFL, MLB and PGA of America, and also customizes apparel with logos and other design touches for corporate clients and organizations.
It uses three DCs to service its business in North America, but the Renton facility does the customized apparel work, using state-of-the art embroidery machines, with the jobs in the embroidery department driving close to 50% of the picking in Renton.
Because of this omnichannel nature, explains Scott Sumpter, Cutter & Buck’s executive director for distribution, enhancement and logistics, the Renton DC has two picking workflows to manage: fulfillment of non-customized apparel shipped direct to consumers who buy online or to retail outlets like fan and pro shops, and another stream of order picking to keep the embroidery production area supplied with the right SKUs and sizes for customized apparel orders.
The company prides itself on its service levels and turnaround time, making the bottleneck in order picking something that had to be addressed. The idea was to find a more automated system so a smaller team of workers could handle order picking, enabling more workers to be trained and shifted to working in embroidery or in value-added services, says Sumpter. Initially, multiple types of automation were on the table as a solution, including mobile robots.
As part of assessing available solutions, the operations and leadership team realized that replacing the pick module with a high-density, robotic automated storage and retrieval system (AS/RS) would provide the biggest leap forward in picking productivity, free up some labor resources to be moved to other areas, and gain more storage capacity. The AutoStore solution (systems integration by Kardex), which went live at the Renton facility in August 2023, was deployed running on Kardex’s FulfillX warehouse execution system (WES) software.
The project included integration between the WES and Cutter & Buck’s warehouse management system (WMS), using integration and orchestration software (SVT Robotics) to tie the systems together to manage processes such as replenishment, order waving and order picking.
The solution, Sumpter adds, is hitting its objectives: during peak season, 12 associates working at the system handled the peak volumes that used to require a staff of 30 in the pick module.
“The solution has absolutely helped us with our picking efficiency,” says Sumpter. “We are not only shipping same day for blank orders consistently with no real challenge, now we’ve been able to reduce our time for embroidery processes, allowing faster turns there. It is because the AutoStore system is getting orders out to shipping, or to embroidery, very quickly.”
Cutter & Buck’s omnichannel strategy—combining online direct-to-consumer sales with sales to retail outlets like golf shops and fan shops, as well as customized apparel orders—has propelled the company’s growth for many years.
With e-commerce volumes increasing rapidly during the pandemic and remaining a growth part of the business, the need for efficient order picking was central to the project. While pick modules involve less travel compared with picking from spread out locations in warehouse aisles, they still involve some walking to get to the correct pick face on each level, pick the line items, and then send that container off to the next level, where more picking might be needed.
By contrast, AutoStore works on a goods-to-person (GTP) principle, with the picker staying stationary at a picker port. The travel in the high-density cube is done by the AutoStore robots, which move on top of the grid to retrieve totes and present them to the picker workstation. There, the associate can follow simple screen instructions and light guidance to accurately pick orders.
The AutoStore isn’t just more time efficient for pickers, says Sumpter, the user interface for the software and visual light cues at the port make the system easy to learn, which helps with on-boarding.
“We can teach people to be productive pickers in about 10 minutes, and a good part of that is safety protocols we go through,” he says. “We are working to get almost everyone, though not our embroidery machine operators, trained on AutoStore so we can move people around fluidly.”
The AutoStore at Renton has 11 ports in total, more than 34,000 bin positions, and 54 robots. It was sized to handle five years of growth, with the ability to increase its output in the future by adding ports, bins or robots to the system.
For direct-to-consumer orders, goods are picked directly into a shipping carton. For the blank apparel that needs to go to embroidery, the goods are picked to tote and brought by carts to a short-term, work-in-process (WIP) holding area near the machines that stages inventory for the next few hours of custom work.
Previous to the AutoStore project, goods for the embroidery work were picked into totes from aisles using orderpickers, and then brought to a three-day WIP storage area. But with the AutoStore now having the capacity to quickly pick goods for both direct-to-consumer and for custom decoration, that three-day WIP area has been converted into a miscellaneous storage area, located right above the DC’s value-added services area on a mezzanine.
During the deployment of the AutoStore, the three-level pick module was kept in operation until after go-live, and then removed entirely and replaced with very narrow aisle (VNA) storage, holding cartons of bagged apparel items for replenishment of the nearby AutoStore.
Over the last several years, Cutter & Buck also adopted VNA racking for reserve storage, to store the SKU range needed to appeal to its customer base. The AutoStore, with its dense cube storage, has 34,000 bins, compared with 24,500 inventory positions in the former pick module.
“AutoStore is incredibly dense storage, so even though it’s a little bit less square footage than the old pick module, we’ve gained storage capacity,” says Sumpter.
The deal for the AutoStore solution was signed in April 2022, with the system going live in August last year. With roughly a year to complete the deployment, Sumpter says it needed a rapid rollout with minimal time spent on software integration.
With limited information technology (IT) staff available at the time the project began, hard coding the interfaces between Cutter & Buck’s WMS (a Manhattan Associates system running on an IBM platform) and the Kardex WES for the AutoStore would have complicated the project timeline, so the company turned to integration software with prebuilt connectors between the solutions.
“That [use of SVT’s platform] was all about project speed,” says Sumpter. “We knew we didn’t have the bandwidth to create the integration ourselves, so we enlisted SVT to help with that.”
The WMS continues to work well for the company and is used to manage functions like receiving, replenishment and shipping, whereas the WES is focused on running the AS/RS processes and order picking. SVT’s software, Sumpter adds, works in the background to pass data and messages between the WMS and WES layers.
Replenishment is typically done using orderpicker trucks to transport cases to a replenishment zone next to the two AutoStore ports used for replenishment. Overall, replenishment is managed by the WMS, communicating with the WES from SVT.
For example, when goods are on their way to a drop zone next to the AutoStore ports used for replenishment, a message passes to the WES that the move is underway and that the goods will soon be ready to be inducted into the system.
Conversely, during order waving, messages are exchanged to see if the AutoStore has sufficient inventory for the orders about to be released, and if it doesn’t, a replenishment gets triggered.
After the AutoStore solution went live, some adjustments were made to the min/max inventory settings to keep enough space in the system for replenishment, says Sumpter. That has now been dialed in to create enough open spaces for ongoing replenishment, but still leverage the storage density of the cube. The end game, Sumpter adds, is to have a smooth flow of order fulfillment throughout the building, with no shortages in the AutoStore, so orders get picked and shipped faster.
Rather than picking to a tote at the AutoStore and then sending order totes to a separate pack out step, the picking for direct-to-consumer or “blank” goods is done directly into shipping cartons.
Associates at the ports erect the cartons from standard sizes, and pack, seal and label each carton, using equipment on a work surface by each port, including bar code thermal printers. Once these direct-to-consumer or other orders for blank goods are packed, the associate slides the carton onto conveyor that will transport the orders to shipping.
For goods being picked for custom embroidery jobs, however, the goods are picked to totes that are moved by cart to a staging area near the embroidery and decoration zone of the DC. To streamline this process and minimize staging space, Sumpter says, the WES was configured so a couple of the AutoStore ports can be dedicated for picking of goods needed in embroidery over the next four-hour window.
“The pickers at those ports don’t have to deal with corrugated or packing—they just pick directly into a tote and when the order is complete, a label prints out that says what machine that tote needs to go to. We run those ports the same hours over two shifts that we run operations in embroidery, so that the flow of goods to embroidery is smooth and stable,” Sumpter says.
The implementation reduced the number of people needed to do order picking during peaks by more than half (12 can now handle the volume that used to take 30), but the overall project driver wasn’t labor reduction, says Sumpter, it was aimed at removing a bottleneck, and being able to reassign people to other areas.
Ultimately, goods-to-person picking is also about speed and cycle time reduction, Sumpter adds, and customer service.
“I think our ability to commit to the fastest service levels that we possibly can has been really positive for the company,” he says. “Our AutoStore deployment has been critical to removing the bottleneck in picking that was previously challenging our ability to hit those high service levels year-round.”
The world of retailing can be unforgiving. You not only need to offer high-quality products that appeal to customer lifestyles, but also fulfill orders with speed and accuracy, across multiple channels. At the end of the day, the solution Cutter & Buck deployed helps to improve service levels for all its channels, not only reduce cost-per-unit picked, Sumpter and Freet agree.
“The big objective was to remove the bottleneck from picking, so we could really make significant leaps forward in our efficiency and therefore capacity throughout our decoration and fulfillment operation,” Freet says. “Enabling a faster turn to our customer while improving our efficiencies was the long-term objective. We have several years of growth still in front of us with the same base system, so I’m confident that we can continue to improve the rest of our operations and keep deliveries super-fast by utilizing the AutoStore system better and better.”