Pick the Best

Periodically, almost every operation is faced with the task of needing to revamp its picking methods. Common goals for a picking process revision are (1) to meet the changing requirements of specific customers; (2) to improve performance by gaining speed or improving efficiency; or (3) to raise the level of accuracy.

Whatever the rationale, to be successful, a revision of picking methods must not only reflect which master is to be served by the change but also exactly how those benefits are to be realized.


The first steps in grappling with a picking method revision are to do some fact-gathering, analysis, and interpretation. These steps fall into several categories. First, what is the actual process today? Second, what will be different in the future, and how much different will it be ? Third, what are the best means of performing the work in the future? Fourth, how fast does the change need to happen? Fifth, what are the steps to get from the existing scenario to the future state? Sixth, what are the implications of these changes, as well as the opportunities they afford, in terms of space, labor staffing, packaging, and other factors?


Though the process is more complex than can be fully addressed here, a few illustrations will shed some light on the value of the approach and process outlined above. Here are a couple of examples of results from the analysis phase:

A large hardware distributor analyzed its line volume by pick area for an existing facility to identify future layout requirements, based on forecast growth by type of product and handling characteristics (see table, p. 23). This in turn feeds fixture and floor space calculations, which drive part of the labor requirement.

Order patterns or profiles can inform both the layout and the picking process, as illustrated by a new facility design project in which a consumer products distributor followed the approach we’ve been discussing. In the best practices phase, the project team determined that shipping used as many as 49 different carton sizes for outbound goods, a fact that was unknown to anyone inside the company. Some simple Pareto analysis indicated that ten of those cartons accommodated 75% of outbound goods. With a little more work, the team reduced the number of cartons from 49 to 21 (43%), generating savings in space, cost to acquire, labor to pack, and inventory carrying costs — all in a single stroke.


In developing picking strategies, it is important to determine the right balance between manual and automated solutions. In its simplest form, automation moves the product to the worker, as opposed to manual operations, in which the worker travels to the product. In many environments, as volume increases, it becomes easier to justify higher levels of automation, because the cost can be offset by greater reductions in the labor needed to perform the work.

Another example will illustrate this point and several others. In this case, the distributor sought to upgrade an existing facility, expecting to optimize performance within five years. The project team, therefore, not only looked at five years of projected data (sales volume, SKU growth, inventory turns, and space requirements), but also developed several options and their related labor implications in years one and five.

In the matrix at right, options one through five combine three alternative picking methods, each of which has different production rates (lines per hour) with three layout options, different equipment investment costs, and different levels of labor savings. All solutions are compared against a current performance level for evaluation purposes.

As is usually the case with improvements of this kind, there are trade-offs. Here, for example, it is worth nothing that both the solution with the greatest labor savings in both year one (65.7%) and year five (51.3%) and the solution with the highest production rates (60-80 LPH) are manual operations. More automation is not always synonymous with improvement, and in this example, the application of automation was not better if one of the objectives was labor savings.

Optimal Picking
1 2 3 4 5 LPH
Pick to cart; pack at shipping (manual) X 60-80
Pick and pack to cart (manual) X X 50-60
Pick/pass; pack at shipping (automated) X X 40-50
Standard with two-level mezzanine X X X
Standard with no mezzanine X
Shipping/receiving at opposite bldg. ends X
EQUIPMENT COST ($ millions) 2.6 2.2 2.5 2.2 2.15
Labor savings — year one 47.3% 65.7% 23.5% 22.5% 44.6%
Labor savings — year five* 37.2% 51.3% 28.4% 27.6% 51.3%
Note: Year-five estimate assumes growth in SKUs and other changes.


Only by living in a cave in the Himalayas can one avoid some discussion of the application of voice and RFID technologies to distribution operations today. Neither is new, but, for different reasons, both are rapidly being deployed in the industry. Voice is no longer constrained by the cost or capacity of its underlying technology and has recently made significant inroads into grocery, office products, and other comparatively low-margin arenas. The most use of voice technology to date has been in the picking function, which usually provides the largest payback. Contrary to some views, this technology need not replace RF or bar code scanning, but it can provide additional incremental benefits, depending on the application. It’s certainly worth looking at.

Total Lines by Pick Area
Bin module 67.4%
Full-case module 8.7%
In-line full case 7.1%
Rack repack: non-conveyable 4.9%
Rack repack: totable 4.6%
Rack full case: non-conveyable 2.1%
Aerosol 1.9%
Heavy (>75 lbs.) 1.8%
Flammable 1.5%

RFID, like voice, has been around and in use in other ways for some time. The major push for it is coming from two major influences, Wal-Mart, the largest consumer goods retailer, and the U.S. Department of Defense, even larger than Wal-Mart.

RFID’s short-term efficacy in the distribution world is not clear, especially since there is, as of now, no single standard and no dominant solution in the marketplace. The cost of delivery for this solution is also among the significant uncertainties at the moment. Speculation on the price of the tags, for instance, ranges from a high of $100 each for high-end, active, read-write versions for high-tech or high-value applications to a low of 50 cents for the simple, passive variety. Some estimates presume that when fully implemented in the industry — probably not anytime soon — tags may cost as little as five cents each.

The advantages of installing this technology in your operations are likely to be greater inventory accuracy, higher fill rates, and reduced labor to handle the product throughout the system. The major disadvantage will be a significant increase in the incremental cost of distribution per unit handled. On the other hand, for suppliers to either the DOD or Wal-Mart, there may be no option. Rather, the optimal strategy may be to determine how to maximize the benefits afforded by the changes that RFID will require.

Ron Hounsell is vice president of software solutions at distribution consulting firm Tom Zosel Associates in Long Grove, IL. He can be reached at rhounsell@tzaconsulting.com or (847) 540-6543.