Bill of material (BOM) management is an extremely complex process. Whether it should be primarily keyed off the functional use of a part or the specific part number itself is a tension that has existed since bills of material were created, when man married an axle to a wheel.
Why do people care whether the BOM is driven by part numbers or the part’s function? The answer is that each has certain advantages. Driving the BOM by actual part numbers is very useful for purchasing and manufacturing and when an engineer is focusing on an individual piece of hardware. However, when the engineer or a product manager moves from one product to another, a function-driven BOM allows him to compare components more easily.
The ability to compare BOMs has huge implications for the ability of engineering to successfully reuse parts. Reuse lowers both the cost of the hardware itself (by reusing tooling and increasing production volumes) while also reducing the engineering time needed to design a new part. Reuse has been the great white hope of the engineering community since time immemorial, but it is very challenging.
Although product lifecycle management software vendors talk about reuse, most companies do a very poor job. Trying to manage reuse without intelligent part numbers and a functional way to look at the BOM is like searching for a single item in a hoarder’s basement that’s jammed full of boxes with no labels. Eventually you give up, go to the store, and buy a new one! That’s one of the big reasons why reuse has struggled in the engineering community.
So how do we balance the tension between these two things? Here is a simple framework for thinking about the BOM that I call Form, Fit, and Function.
By function, I mean what the part is actually doing, what its purpose in life is. By form, I mean the specific piece of hardware or part that we’re dealing with. Figure 1 is a simple example of how function and form relate to each other. Typically, the superset will be function. The function in this case is to provide rear vision to a driver of a vehicle. The function is the result of the part operating correctly. The part itself is the form that “delivers” that function. In this case, we might provide rear vision by designing a mirror, a camera, or some sort of sensor.
The third part of the framework is the fit of the part. However, by fit I don’t mean the actual geometric tolerances or real estate that the part occupies. Rather, what are the unique attributes of the part that make it “fit” the function, i.e., accomplish the function?”
Figure 2 shows another example of a functionally controlled BOM. In this case, it deals with the propulsion of a vehicle, or perhaps a jet aircraft. We show propulsion as a top level Function, along with sub-levels of the powerplant, how the power is transferred, and the cooling system.
The coolant is the Form that satisfies the Function of the cooling system. We assign an intelligent part number for the specific coolant. One might ask, “When do I move from a Function to a part number (Form)?” The answer to that is, whenever it is convenient. It will take a little bit of time for your team to develop a functional BOM that has a manageable level of hierarchy in it.
There is a many-to-many relationship between the Function and the Form (part). Depending on how far down the Function tree we go, we may need to attach more than one Form (part) to satisfy a function. On the other hand, if our functional tree is deeper, on certain products there may be one assembly (Form) that satisfies more than one Function on the functional tree.
Moving to Fit, we see that each Form may have multiple attributes (ways of fitting the function). The coolant “fits” our functional need with the attributes that it has. In general, there will be a one-to-many relationship between Form (part) and Fit (attributes), respectively.
Form, Fit, and Function is a simple way to look at how we structure our products. It lends itself well to reuse of parts, but also for the reuse of work break-down structures that are used in aerospace and defense.
Some companies are now working on constructing skeleton “starter” or “universal” BOMs that they reuse for each new product. The idea is to start with a generic BOM and then add and delete to match the needs of a new product. The goal is for the company to have one universal BOM or one for each unique product group.
This is a great idea in theory, but it’s not trivial to execute. To do this in your own company will likely take your engineering team, product management, and a consultant a year or more to find a structure that suits your needs. However, once this is done, the speed and reuse advantages should be significant. Hopefully, the Form, Fit, and Function BOM framework will give you a simple way to think about it!
Eric Hiller is the managing partner of Hiller Associates, an operational and strategic consultancy specializing in product cost management, financial modeling, operations, business planning, software product management, and product development.