In an earlier issue, I offered a five-step test to determine if you're ready for automation. The steps started with defining your materials handling problem and finished with determining if you have the skills and discipline to manage an automated operation. If you found yourself in pretty good shape, then now is the time to think about choosing the right level of automation.

But before you get going, I have some thoughts for you. Recently, I was asked to recall the most and least successful applications of automation in a warehousing environment that I had seen. The reflection was an instructional process on what to do, and what to avoid when developing an automation project.

Manage your ambitions. The biggest disappointment comes, of course, when we have not done due diligence with the data (see April 2003 column - Delving into the data), and we actually solve the wrong problem. Beyond that, the most troubled systems that I can recall are those that were too ambitious for their budgets.

In an attempt to be everything to everybody, the designs became progressively more complicated. Complete integration of diverse technologies is difficult, but not impossible.

As the system grows, the diminishing savings from the incremental functionality puts cost pressure on every element of the system. Only the wildest optimism regarding cycle times, up-time and system balance will provide calculations that meet the throughput requirements.

Keep the flow simple. The most successful systems seem to be those that simplify the flow.

In fact, the most productive orderpicking system that I have seen is a simple belt conveyor, moving order cartons slowly but continuously past zones of light-directed picking from flow rack. If a carton reached the end of a zone before its picks were completed, the conveyor stopped until they were finished, and then restarted. At lunchtime, several products were re-assigned from the zones that stopped the conveyor most frequently to those that were more lightly loaded in order to provide a better balance for that day's order mix.

In successful systems, individual components are designed to simply cope with the extremes. Peak flow rates are managed without strain. Variations in carton size and weight are handled reliably without jamming. Yet, the system must withstand the abuse that comes with years of service. Remember, the system needs to keep running far beyond its payback period to be a worthwhile investment.

It's all about the money. Don't get trapped into supporting the return on investment with wishful project budgets. You will be in for a year, or two of misery—and, surely, disappointed at the end. Estimate the savings conservatively, and the costs at the high end of the range. Allow for contingencies so that unexpected price increases don't cause you to compromise on vital functionality. This may make it harder to sell some projects, but it's better to stop before the first dollar is spent than to set off on an impossible journey.

So, there it is! Finding the sweet spot of success not so difficult after all. All you have to do is keep the concept simple, the design robust and the budget adequate. You'll be glad that you did.