By Ian Hobkirk
Managing Director of Commonwealth Supply Chain Advisors
In our work, we are constantly seeing examples of two companies with cart-based or conveyor-based distribution center picking systems – often in the same industry and with similar order profiles – but with vastly different levels of labor efficiency. The lower-performing operation invariably believes that there is nothing they can do to change the situation, and that without a massive investment in automation, their distribution center pick rates must remain as they are.
However, when we start to really look under the hood of the distribution center operation – by close observation and data analysis – a few common root causes of the under-performance usually emerge, which are almost always easier to address than the subject companies realize. Here are three causes which often account for lower-than-necessary pick rates:
- Poor SKU density: not enough SKUs are slotted per walking-foot
- Poor line density: not enough “pick lines per walking foot” occur in a typical pick route
- Ineffective pick methodology: the use of basic pick methods such as discrete order picking or single zone order picking
For each of these root-cause issues, the result is the same: for every linear foot of travel by picker, there are too few picks performed. Below are a few basic ways to rectify this situation and elevate distribution center pick rates:
Store the right supply: Many companies hold too large of a supply of product in forward pick, not wanting to expend the labor to replenish this area too often. While replenishment labor is a trade-off that must be weighed against any improvement in pick rates, it is one that is often worth making, as replenishment is usually performed in more efficient, larger quantities than the piece-meal picking process. Many companies with cart or conveyor-based pick systems use a replenishment interval in the two to four week range.
Use the right storage medium: Many companies have pick modules that consist of too much carton flow rack and not enough smaller storage mediums like shelving. As a result, pickers can’t access enough SKUs in a given pick path to operate efficiently.
Use walk-back aisles properly: Many shelving-based pick modules have design flaws that make it difficult to pick quickly. Often, aisles are too narrow to allow picking carts to be taken down the aisle, or to allow two carts to pass in the same aisle. As a result, workers must either limit their picks in a given aisle to just one line item, or they must grab multiple items and sort them out when they walk back to the cart, leading to errors. Wider aisles will of course help to rectify this problem. However, if this is not an option, then using the WMS to reconfirm items as the put process takes place can help allow multiple lines to be picked in a visit to an aisle without sacrificing accuracy. Companies using voice-directed picking may encounter some limitations with regards to re-identifying and re-confirming picks in this mode of cart picking. It is often more effective to re-scan the items at the cart than to try to read a lengthy product ID back into the WMS to re-confirm.
Zone pick effectively: When a pick-density analysis is performed on under-performing operations, it often reveals a degradation in efficiency due to the combination of slotting and pick methodology in use. If items are slotted by velocity, and picked in a single zone, each pick route generally tends to start out very efficiently, as picks in the high-velocity area are made. Often though, pick efficiency sharply degrades as the picks are made in the slower-velocity zones with less pick-line density. One way to rectify this is to split the picking process into a multi-zone pick. When the slower moving product is actually isolated in a separate zone, with a dedicated pick route, then productivity can sharply increase. This is, of course, not feasible in all operations. But, it is often possible to create a few waves each day that include a large number of orders that all contain picks in the slow-velocity zones. This tactic boosts the pick density in this zone to higher levels than normal. When the picks are completed, they are then either passed to the high-velocity zones for the remaining picks, or consolidated with those picks which have already been made on other zones.
None of these solutions involve a fundamental change to the overall picking strategy. Implementing these tactics may, however necessitate changes to the WMS system which should be thoroughly investigated. If the changes to the WMS are deemed too costly to be made at the present time, it may be worth considering solutions like voice-directed picking. Voice software may allow some basic changes to pick methodology using standard functionality and will not require code changes to the WMS (see the cautionary note regarding voice picking in walk-back aisles above).
Companies that are struggling to raise distribution center pick rates would do well to investigate these four solutions as ways to reduce labor costs and improve cycle times. Commonwealth’s new whitepaper, Confidently Committing to a DC Design – When Demand is Unpredictable Part II provides a more in depth look at how to select a picking strategy and why it is so important.