DHL Supply Chain, a leading provider of warehousing services and part of Deutsche Post DHL Group, the world’s largest logistics company, recently announced plans to deploy emerging technologies in 350 of its 430 facilities in North America. This is a $300 million investment. DHL Supply Chain released a press release on this, but my colleague Chris Cunnane and I had an opportunity to dive deeper into this. We had a conversation with Scott Sureddin, Jim Gehr, and Adrian Kumar – North American CEO, President Retail, and Vice President of Solutions Design, respectively.
The press release mentioned investments in robotics, augmented reality, robotics process automation, IoT and DHL’s visibility solution – MySupplyChain. “Selected technologies will vary by customer needs, based on the outcomes of research and pilot programs completed by DHL’s internal innovation teams and collaboration with dozens of external innovators.” All well and good, but Chris and I wanted to know which of these solutions was showing the best initial success, and which vendors in different areas they were most entrenched with.
I was surprised we did not spend most of the time talking about robots. Indeed, one very interesting part of the conversation surrounded how they were getting more out of well-established software systems. Already in active deployment for many years are warehouse management systems (WMSs) from Manhattan Associates and JDA. DHL believes they have deployed both supplier’s systems at more sites than any of their other customers. Many of the new technology solutions DHL is investing in are focused on automating the labor-intensive piece picking requirements driven by ecommerce. Manhattan is their solution for ecommerce. Manhattan, like all major suppliers, have new releases every year. “Many of the enhancements are DHL driven,” Mr. Sreddin asserted. “We have lots of influence on the next release.”
That makes sense. Besides being a big customer, DHL is also a sophisticated customer. Historically, many WMS solutions were not designed for piece picking. They dropped orders to the warehouse in big batches. This intelligent grouping of orders worked well for case and pallet picking, but “it does not work well in ecommerce,” Mr. Sureddin explained. The software needs to drop orders more continuously. The system needs to ask itself, “What order should we release that will not cause an additional bottleneck? The workflow must allow workers to be more fully utilized.”
Right now, two-day shipping for ecommerce orders is the norm. Companies that can get to one day shipping are at a competitive advantage, Mr. Gehr pointed out. More intelligently dropping orders to the floor can improve picking productivity by 25 percent. This in turn can allow companies to push the cut off time for shipping back several hours and ship a greater percentage of the orders that drop on a particular day. Manhattan has done interesting product development in this area.
But the 25 percent gains DHL is getting in productivity surrounding optimized order drops to the floor go beyond the new capabilities in the WMS. DHL is using simulation to help them set the proper decision rules for when and how to drop work orders to the floor. They are simulating the productivity in both picking and packing and seeing how the bottleneck changes based on the order drop rules.
New systems are also being overlaid on these existing warehouse management systems to aid in driving better productivity and throughput. Mr. Kumar talked about the material investments in data lakes that DHL has made. “We hire 400 college graduates every year” including many young data analysts that are “aggressively working” with this technology. They are moving up the continuum from better understanding what has happened based on historical data, to greater use of current data to improve real time operations, toward their target of more fully using predictive analytics.
What does predictive analytics mean in this context? DHL is working to anticipate how many orders will drop to each warehouse every day of year. This is based both on historical data, but also on pulling in external data from the marketplace. And over time, the “systems start to learn.”
Mr. Gehr added that when the company was not seeing the volume of work they were expecting on a day, or even a certain hour of the day, they reach out ask their customers to follow-up. “It could be issues have popped up. We want to run a preventive supply chain.” It could be the orders are out there, but there are issues with inbound transportation, or something is wrong with the ecommerce system and orders are not dropping. 60 to 70 percent of our costs are labor.” There can be 2000 people working in an ecommerce warehouse; you want them all effectively deployed.
Our discussion also included robotics. They are using automation for palletization and depalletization, robotic process automation for automating clerical activities like freight processing, a new generation of automated guided vehicles that can move pallets from different origins to different locations and don’t require changes to the infrastructure, goods to man automation, and of course autonomous mobile robots (AMRs).
When it comes to AMRs, they don’t believe there is a one size fits all. The best system can depend on SKU velocity (how frequently a product sells), the layout of the facility, and other factors. They have announced they are working with Locus Robotics at more than one facility. But in this emerging area of automation, no vendor would have the scale to meet their needs if they decided they needed 10,000 robots. Robots are currently deployed at 85 of their facilities. In some cases, they have seen productivity gains of 200 percent. “But,” Mr. Gehr points out, “it is not plug and play.”
In conclusion, it is a common observation that improvement is usually based on people, process and technology. Mr. Gehr admitted that some reeducation is required, particularly for their managers. “What we did in the past was great, but there is a different and better way to run a distribution center. People need to accept that things are different. And we still need our workforce to apply their existing knowledge.” But the company has 35,000 associates. “We need to continue to invest and to be an industry leader in digitalization. This is important if we want to be an employer of choice.”