When we begin to analyze information flow in terms of lean, we must continually ask two questions: one, how can this information be supplied more efficiently and two, what information would make this particular process or operation more efficient. In other cases we need to ask what information is being created and supplied that has no actual value and, thus, is wasteful in and of itself.

 

The first challenge to efficient information management is the management of the process itself. All too often the person needing the information has limited knowledge of the technical aspects of information management and the person responsible for supplying that information has limited knowledge of the needs and application. The net result all too often is information of limited value supplied in an inefficient and ineffective format.

 

Without exception, the one characteristic that I see separating good companies from great companies is their ability to manage information flow. The great companies gather input and process information, which adds value in supporting their manufacturing process. Most often the volume of this information is huge, but it all is designed to support the flow of materials on the plant floor with as little disruption as possible, thus minimizing waste. 

 

Another common thread which I see in these great companies is information technology (IT) people who intimately understand the manufacturing process — and this is a key element. 

 

One of two things must exist to maximize the effectiveness of IT managers. Either they must be an information management specialist with an intimate understanding of the manufacturing process or they must be a manufacturing specialist with in depth knowledge of information management — it tends to be a rare bred.

 

The end result is an individual who understands what information can reduce waste and add value to the manufacturing process and knows how to deliver that information in the most efficient cost effective way.

 

If you happen to wear both hats, it is most likely that your manufacturing hat fits best; you need to have or develop a working knowledge of information management to be able to understand both the capabilities and the limitations. One of the most valuable courses I took in college was computer programming. Despite the fact that I date myself when I say that was Fortran, that experience has proven invaluable to me. 

 

With those thoughts of “who” in mind, let’s look at what it is we want to do with this information. Some of the key elements in an efficient manufacturing process are: not to produce anything for which we do not have an order; to produce that order on a just in time basis; and to never begin producing an order until we know everything to complete it is available. In cases were multiple components are produced internally to go into a final assembly, we need to be assured that all these individual components are where they need to be — and possibly more importantly have in place a method to detect and recover from a missing or damaged component.

 

Equally important is how we manage that information transfer across the entire operation. We do not want to enter data manually as it moves through the process. Manual data entry only increases the chance for error and represents waste. Optimally, the system will be self-checking, thus verifying good information along the way.

 

By whatever method you get orders into the system, through customers, dealers or your own office people — when that order is entered it should be automatically verified as a clean order. That means every item on that order is complete and will work with other elements without conflict. Even in custom millwork and store fixtures, elements of this self-checking feature can be utilized.

 

When an order is released to the shop floor for production, there should be automatic system verification that order can be produced; that every component and part is or will be available to complete the manufacturing process.

 

Labeling of individual parts might seem to be a waste to some, but I find it to be just the opposite. A label affixed to each part at the saw, with the proper information either contained within or retrievable from the barcode, ensures a tremendous return on investment.

 

Normally the label will carry both readable text and a unique barcode. The form of that label is immaterial; it can be any suitable format or even a Radio Frequency Identification (RFID) chips — the important thing is the information contained and how that information is utilized in the process.

 

One of the benefits of individual part labels is part tracking. Scanning the label at each work center allows locating any part in the system; it also offers a checking capability. In the event a part goes missing, the system can be designed to recognize that, and to reorder a replacement part. In the event a part is damaged in the process an operator can have the ability to immediately reorder a replacement.

 

The same labels can be used to set up all the production machines along the way. This allows associating a particular part number to a machine configuration or program. Most often these programs will be retained on a central computer remote from the production floor which sends down the correct machine program when presented with the label information by a scan at the machine. The time to do this is normally measured in milliseconds.

 

The storage of all machine programs in a central data base, in a generic format, allows the same information (and part) to be run on different machines since post processing is being done at each individual machine. Another huge advantage to this central data base is that when programs are revised, only one program is revised not multiple programs stored at the individual machines.

 

Where assembly operations are involved, these same component labels can be utilized to generate assembled ID labels and shipping labels at the point of use. As those items are loaded on trucks, the carrier information can be added and invoicing can be automatically triggered. This same information can and should be utilized for inventory management and reordering.

 

Ultimately, the goal should be an enterprise-wide information system that maximizes efficiency, minimizes data entry and works across the entire organization transparently and seamlessly.

 

That almost sounds easy — but it isn’t and it takes a lot of planning, many many hours in execution and, yes, even a few grey hairs to accomplish. It is not necessary to overlay an enterprise level information system at one time — I would say it isn’t even advisable. The element that needs to be most focused on is a system that will ultimately provide the information you need in a format needed, and minimize human intervention with respect to data entry. A good system will allow the laying of a foundation or spin that will then support individual applications as they are added.

 

Success will come once you have defined in detail the ultimate goal of your information system; then the individual steps to accomplish that goal can follow. Avoid what seems to be a common error of putting in place “point solutions” to problems that ultimately cannot support an enterprise solution — establish your goal up front, and then work to that goal.