By John Diebold,
Director of Supply Chain Consulting, Commonwealth Supply Chain Advisors
I have been around carousel systems most of my career. For those who don’t know, horizontal carousels are, simply put, one of the most popular order picking solutions with thousands of machines in operation around the world. Carousel systems are reliable, fairly budget-friendly and capable of providing dramatic increases in throughput and accuracy over manual picking operations. And yet, despite their countless benefits, I continue to meet clients who are in the process of having carousel systems removed or letting them gather dust in a far off corner of the operation.
I have designed and managed many successful projects involving carousel systems of all shapes and sizes. I’ve also seen a few carousel projects that never lived up to their expectations. Sometimes the technology is a poor fit, but more often than not the success or failure of a horizontal carousel project can be attributed to a short list of culprits:
- Storing the Wrong Product in the Machines
- Improper Use of Technology
- Political or Cultural Reasons
- Poor Slotting
Storing the Wrong Product
The primary goal in selecting carousel candidates is maximizing pick density while reducing replenishment transactions. Items designated for storage in carousels should strike a balance between size and usage. If the pendulum swings too far in either direction, the amount of time an operator spends replenishing and waiting will increase.
Improper Use of Technology
Horizontal carousel systems spawned from the garment conveyor systems you find at most dry cleaning operations. These carousel systems typically operate using a simple location keypad and do a good job of providing one or two picks per minute. Their horizontal cousins found in distribution operations worldwide use optimization algorithms, pick lights and supplemental equipment to regularly generate 300 plus order lines per hour. Unfortunately, a growing number of carousel systems are pieced together with used equipment, fashioned with outdated controls and little attention to software. These, as you would expect, often generate throughput comparable to a dry cleaner instead of a world-class distribution system.
Sometimes, operations go the other way and overburden the system with bleeding edge technology and layers of homegrown software. Often these components increase downtime and give the illusion that the typically reliable carousels are “always down”. So when your carousel provider offers to install a 3D holographic projector to display the likeness of the next part to be picked, I suggest you opt to keep it simple and stick to proven solutions.
Political or Cultural Concerns
About two years ago, I visited a client who recently installed a pod of three carousels and was extremely disappointed with their effectiveness. It took less than a minute to realize that the operators had simply filled each carousel shelf with a single random SKU. This resulted in a fraction of the storage capacity and terrible throughput rates. Discussions with the customer revealed that the carousels were the owner’s decision, and the machines were loaded by a skeleton night crew without any training or knowledge of the slotting plan for the team.
A great number of carousel missteps are simply a result of miscommunication and poor planning. It is surprisingly common for a horizontal carousel system to be purchased by someone high up in a company and thrust upon a warehouse team with little or no warning. Without some basic guidance on the principles of what goes into a successful implementation, the team is left to flounder and ultimately sour on the idea of carousels in general.
Slotting
Slotting is the art of directing product to a properly sized bin in a velocity appropriate position within the operation. I purposely reserved this item for last, as I believe the three principles above must be taken care of before the true benefits of slotting can be realized. Some of my clients and colleagues will laugh when I say that “slotting is overrated,” but in my experience, when a carousel system doesn’t stumble on any of the three prior bullets, you will likely have a system that can outperform a decent manual picking operation. [Technically speaking, item #1 (Storing the Wrong Product) can be considered an element of slotting, but it was separated to highlight its significance]
Some of the keys to a successful carousel slotting strategy include:
1. Organize your product so that the majority of activity takes place at the height at which your operator is most comfortable working. It is important to focus on pick density, as often a number of smaller medium velocity SKUs account for more activity than a larger high velocity SKU.
2. It is extremely important to balance the activity between all the carousels being serviced by a single operator. This can be accomplished quite easily by sorting the list of items to be stored and assigning them to the machines in a round robin fashion. An unbalanced system will often result in an operator continually waiting for the same carousel instead of rotating between units.
3. Maintain a strong focus on replenishment. Locations should typically be sized to accommodate a couple weeks’ worth of picks. It is also good practice to ensure the location can store quantities proportional to the handling units, to avoid constantly topping off with partial case quantities.
4. Probably the most important rule is to, once again, keep it simple. Most carousel slotting plans can be created using a handful of location types and 1-3 velocity codes. Having twenty different location sizes will allow you to find the perfect size location, but the additional complexity is rarely worth the trade-off.
Implementing a successful horizontal carousel system is sometimes a journey. Unlike many alternative solutions, they require quite a bit of planning and discipline to get the most out of them. Poor planning can lead to issues that can quickly snowball and, unlike manual storage systems, there is rarely a simple workaround. This state of “working without a net” probably leads to a number of systems being prematurely shut down. Some of these systems may be genuine poor fits, but it might be worth giving them a second look to see if some simple changes might make a world of difference.