Two weeks ago, our colleague Jasper Veurink posted his blog about warehouse automation. More than half of the Top-10 food retailers in the Benelux have now automated parts of their warehousing operations or are working on it. These are exciting projects and the automation industry is booming: companies such as Vanderlande, Dematic and Witron are growing rapidly. On the other hand, according to Peerless Research Group, 42% of all the warehouses are still mainly manual and 51% is a mix of automated and manual processes. The shortage of labor is increasing the need to rethink the labor-intensive warehousing operations. But what if you are not ready to make that big investment, or you feel that there should be ways to optimize that manual way-of working? And without doing this first, will you ever harvest all benefits from the automation investment?
Therefore, we share a pragmatic Lean methodology in conventional warehousing which you can use to check where you stand in your conventional warehouse. If you would invest time in optimization according to the methodology, you will be in a much better position to move forward. And you can always revisit your business case for automation and see what the newly reached productivity numbers have done to your ROI. The challenge with this is that the analysis is the easy part. Making the actual improvement is the hard part.
If you want to learn more about the theory of Lean, we recommend that you read “The machine that changed the world” (James P. Womack, 1990), “The Toyota Way” (Jeffrey Liker, 2004) and “Lean thinking” (James P. Womack and Daniel T. Jones, 1996). These 3 books together will give you a thorough background of what Lean means, what it does and why. Our ideas are based on these principles, but translated into a pragmatic approach.
Our approach is based on the so-called 5-touch warehouse process. This is the operational model for a warehouse with a buffer stock and picking operation. In 5-touch-warehousing goods are received (1), put-away in storage (stored; 2), replenished to a pick location (3), picked (4) and finally loaded (5). Hence the five touches. Warehouses that apply other flow types can apply less or more steps in their processes or certain steps such as put-away and replenishment can be explained differently. In the end, also these warehouses can still follow the logic.
Lean warehousing philosophy: remove the 7 forms of waste
The Lean philosophy tells us that we should remove waste (or Muda in Japanese). Waste is generally described as activities that do “not add value to the customer” and comes in 7 ways:
– Excess Inventory (too much stock)
– Waiting (too much time between process steps or stops within the process)
– Movement (movement (of the operator) within a process)
– Defects (errors, mistakes, re-work needed)
– Transport (of goods between process steps)
– Over-processing (too much handling/ steps for the same result)
– Over-production (too much work being done)
Analysis: the easy part
If we look for the 7 wastes in the 5 touches, we have 35 areas to investigate. This is the easy part of your project. Simple put the 5 processes under the microscope and check if you see excessive stock, operators or goods waiting, movement that can be simplified, and so on. Start at reception and work your way through the warehouse.
Most apparent examples: waiting
Since picking can be 40-60% of the work and is therefore the most labor-intensive activity in the warehouse (Tompkins et al. 2003) the most apparent example of improvement potential is shown when the operators in this team are waiting. A first action is to check and improve the forecasting of the work (total volume per day and fluctuations across the day), and planning of the workforce. Secondly, the picking process should be prioritized over the other processes. For example, the replenishment process should be organized in such a way, that a picker is never waiting for stock in the pick location. This can be done by letting the WMS prioritize the replenishment assignments and planning enough resources in this team. In the case that goods are not available in bulk storage for replenishment, the order line(s) should be removed from the picking assignment. Thirdly, congestion between pickers and other operators should be avoided by workload planning across the processes.
Movement and transport in picking
The second priority of Lean optimizations should be in the picking movement and transportation. Again, because picking is most labor intensive, the routing of the picker before, during and after the picking should be shortened as much as possible. This means that both the routing as well as the length of the pick-run should be well thought-through (routing strategies will differ per type of operation). This has also implications for the type of pick locations, the start- and end-point of the pick-run, the drop-off points, printers, etc. During the picking, taking the goods from the pick face should be made as easy and comfortable as possible for the picker. Finally, an important consideration is that if the picker could take goods for multiple orders and/ or customers during one pick run, he or she can benefit the most from the distance that must be travelled anyway.
As third priority, we would like to mention excess inventory across the warehouse. As a start, excess inventory in the reception area (more incoming goods during a certain period than the warehouse can properly absorb) will create a snowball effect of inefficiencies. It will become more and more complex to perform solid checks on the incoming goods and perform proper put-away. When the checking comes under pressure, more mistakes will be made (defects) and corrective actions are needed (over-processing). It also will be more difficult to maneuver with the goods, making the following processes less productive.
Following upon reception, excess inventory in storage limits the possibility to find the optimal location for goods. Therefore, it means that goods are stored far away from the reception, picking and loading area, creating excess transportation. It can also force the warehouse team to relocate goods from time to time, because the space is needed for other goods, which implies a form of over-processing.
Finally, excess inventory in the loading area finally is a signal that the integrated throughput planning in the warehousing is not functioning well. This means that goods arrive too early in the loading area, blocking docks for a longer period than strictly necessary. So-called “Just-in-time” planning is not working properly. The indirect effect is that other loads are positioned at docks that could be far away from the picking zone, creating excessive transport (similar to what happens in the reception area). What also could happen is that orders are changed or even cancelled during the waiting, which means they must be put back into storage or partially repicked. This could have been avoided by starting at a later moment.
How to improve? The hard part
After analysis of the 35 possible improvement areas, the hard part of implementation and change starts. The analysis can be done rather isolated from the operation but for the improvement you will need the entire team. This will be an organizational change project that will potentially encounter all difficulties that such projects do. On top, there are additional challenges: logistics is predominantly a slim-margin business, and still not always regarded as a strategic advantage (despite all efforts from the supply chain community to explain this). Therefore, managers are reluctant to add specialized resources and budget to drive the project. This means that the project will have to be done by the team itself with limited or junior resources. Additionally, the industry has a strong tendency to be happy enough with “good” rather than “great” and Jim Collins has already taught us that “good is the biggest enemy of great”. Finally, improvement projects must be a top-to-bottom & bottom-to-top thing and can (depending on the maturity state) last multiple years. This means that strong and continuous support from the top is needed for a long period, trusting that this will pay off.
Besides the actual improvement ideas, each improvement program should start with measuring the current KPI-scores. This measurement starts with the service level KPIs such as on-time-in-full (OTIF) and accuracies on picking and stock; the quality KPIs. The quality KPIs indicate how well your operation is running. Second, measure the productivity KPIs, generally explained as output divided by input (how many working hours do you need to receive, put-away, replenish, pick and load the goods). Start with the critical processes like picking and replenishment and display the KPIs as openly and clearly as you can so that nobody will be able to miss them. Also discuss these KPIs every day with the entire team during short and intensive stand ups (do not keep this to the management team alone, but include all operators). If you don’t measure the KPIs from early on, you will not be able to explain and/ or understand the impact of your program.
Once the as-is performance is clear, idea-generation is the first step in continuous improvement (Kaizen in Japanese). Each team member should be allowed and motivated to come forward with ideas (big or small). This engagement is probably the hardest part of the implementation. But the more the team itself comes forward with ideas, the easier the implementation will be. An improvement program leader should provide feedback on all ideas openly: which one’s will be implemented and which will not (yet) and what is the status. The bottom-up idea gathering is critical to engage the entire team on the performance. Once the ideas come forward in a more proactive way, work with an open idea board and incentivize the best ideas with recognition and/ or small rewards. Make sure the top management is connected to the shop floor and pro-actively approaches all team members for suggestion. These walks on the shop floor (so called Gemba-walks) ensure that everybody understands working on the performance is a team effort.
There are multiple ways to select priority improvement areas. A first step can be to consider the impact and complexity of each optimization idea and select the high-impact low-complexity ones to start with. This will most likely make you start at the picking process. If you will find that the other processes (specifically replenishment) are important enablers of successful picking, you might want to start there. Finally, it can also be an option to follow the goods, starting with reception and ending at the loading process.
Paul Hehenkamp is managing consultant supply chain management at RGP. He can be reached at firstname.lastname@example.org
 Publicly known to be working with/ on partial or full automation: Albert Heijn, Colruyt, Delhaize, Jumbo, Lidl, Plus. Not publicly known to be working with/ on automation: Aldi, Carrefour, Detailresult, Sligro