In this technology-dependent world, we often assume that automation must be the answer to every problem. Not so in warehousing. Most warehouses and distribution centers have shied away from large-scale or total automation due to its cost and complexity.

Yet warehouse automation, when judiciously applied, clearly provides significant benefits. Some consultants, therefore, are extolling the virtues of partial mechanization, a practice referred to as "islands of automation." This involves selectively automating certain functions, such as receiving or put-away. Done properly, the resulting reduction in labor costs and increased efficiency is sufficient to justify the cost of the new materials handling system.

Setting up islands of automation isn't for everyone, but for companies that carry out a successful island-style implementation, the rewards can be surprisingly large. "The impact has a lot to do with the size of the company," says Darrell Krasoski, principal at Tompkins Associates in Orlando, Fla. "But if done effectively, you can see a 10-percent to 15-percent jump in productivity."

One company that chose the "island" approach is third-party logistics provider Halo Distribution LLC, which specializes in picking, packing, and shipping products for online retailers. Shipping volumes in that business are frequently subject to peaks and valleys, so Halo wanted more flexibility in its 40,000 square-foot distribution center in Louisville, Ky.

The key to getting that flexibility was creating an island of automation for order picking. It consists of two 20-carrier Remstar horizontal carousels, each 25 feet long with pick-to-light control for broken-carton picks.

The carousels are teamed with a pick-to-light flow-rack system for full-carton picks, and a pick-to-light shipping carton batch table. Fully integrated software provided by Remstar drives the automation. Halo's server receives orders from clients, and the software manages fulfillment by controlled induction of batched orders.

That system allows a single operator to pick 450 to 500 lines an hour. When things get busy, adding one more operator can boost that number to 700 lines per hour, depending on the order profiles. (The average shipment is two lines totaling five items per carton.) "The results have exceeded our expectations," says General Manager Stu Gilray. "We have seen significant improvement in system stability and pick speed since installation."

The "island" approach to automation was the right solution for Halo for a couple of reasons. First, the carousel/pick-to-light setup provided the greatest productivity for the lowest cost. Its design flexibility also lends itself well to managing future volume increases. The DC is set up to accept two additional carousels and two additional flow racks at floor level, and all four carousels can be lengthened to 40 carriers if needed. Should order volume increase so much that a larger DC is required, the carousels could be moved quickly and easily.

That kind of flexibility has proved to be crucial in an e-commerce environment, says Krasoski. "Before e-commerce, you could plan further out," he says. "Today we're all more gun-shy about the accuracy of our 'crystal balls'—it's hard to know what a five-year forecast will be. That's the advantage of taking automation one step at a time, rather than all at once."

The island approach to automation is more common than not, notes Don Derewecki, executive vice president at Gross & Associates, the Woodbridge, N.J., consulting firm. That's because financial issues play a major role in the decision to automate just a portion of a DC.

"It's very difficult for a company to come up with the justification to automate everything," Derewecki says. "Capital is always a limitation, so you have to prioritize where you'll make that investment."

That investment can end up being much bigger than expected because automation has become so complex in terms of information and mechanical flow, says Ken Thouvenot, a vice president at Alvey Systems, a St. Louis, Mo.-based manufacturer of materials handling equipment. "There's a risk in implementing too much automation because it requires so much time, energy, and resources," he says.

Few companies, moreover, have information technology (IT) departments with the manpower to support an all-out automation effort. Even companies that do have the finances and IT support may find that such an approach isn't desirable. "Sometimes companies have the technical capability to automate an entire facility, but if they do so, they lose flexibility and reliability," Derewicki says.

With those issues in mind, many warehouse managers choose to begin with an island of automation when they need to test a new concept or process. Once they've achieved success with the first implementation, they can proceed to the next step.

"Companies want to show that there's a financial justification for automating a specific area of the DC," says Thouvenot. "They start in one area, prove the concept, and then move on."

How can you determine if islands of automation are right for your DC? First of all, be sure that your baseline data is accurate and that you have an accurate picture of the problem to be solved. "It's not always the equipment," Krasoski says. "It might be the training or processes you've put into place." If all the information about your processes is correct, then the next step is to ensure that existing technology, such as your WMS, is meeting your current needs.

From there, Derewecki recommends, review your overall operation and determine which factors are driving your need for automation, regardless of how small a scale is involved. "The classic approach is to look at your labor-intensive functions," he says. "It's common also to look at those areas where it's tough to meet customer requirements and quality levels using just manual labor."

Thouvenot agrees with that approach. "Focus on your labor-intensive areas," he says. "You want to automate for labor reductions and to lower your inventory requirements."

Other areas in the DC to study for potential island-style automation include inbound operations, storage and retrieval, picking, print-and-apply areas, and outbound operations.

Before automating any individual or group of functions, the experts say, be sure to accurately assess costs and consider alternatives. "Do an analysis on several factors, including flexibility and performance," suggests Derewicki. "How would hiring more people compare to automating?"

It's equally important to think through the potential impact of any change, Krasoski says. "Look at the big picture and see how making one change affects the rest of the operation, along with the problem itself," he says.

Finally, don't automatically assume that automation is required. "Too many companies look at their problem areas and spend money to automate the area they think is their stumbling block," he says. "It's much better to have a corporate strategy and take a holistic approach," Krasoski believes.

In other words, do your homework carefully. Although automating one area of your warehouse might be the solution to your productivity problems, it isn't always the only and best way to go. The bottom line, says Krasoski: "Automating for automating's sake won't solve any problem, no matter how much money you invest."