Although today you may be a “board footer,” or a “material times factor guy” or even a “time and motion person,” at sometime in the future you will be pricing your trusses and wall panels using a completely different system then you are today. I can I be so certain of this? It’s because there is a way to price components that simply makes too much sense not to use, and the tools to enable you to use this method either already are, or shortly will be, available to you.
What computers are good at
Beginning in the 1980’s running each individual truss on a computer replaced referencing a standard design in a manual. It did not take long to realize that the computer’s ability to calculate the truss’s material content (BF) or cost ($), (while checking the structural integrity) was too valuable and easy NOT to use. This spawned The Big Three pricing methods:
Board Foot x Factor – Flexible in that it can be used regardless of truss type (girder, attic,) simple in that you are dealing with one computer-generated number (BF) and one variable (the “Factor.”) Limited in that the Factor must take into account a lumber cost index, labor, markup and overhead.
Material Cost x Factor – Same idea – one computer generated number, one variable. Board footage and material cost tell the estimator nothing about the complexity of the truss, so either the complexity is ignored or has to be subjectively applied by looking at the profile of each truss.
Time and Motion – This is my shorthand name for any labor estimating formula composed of setup factors and “run” factors. It uses any number of calculable items about the truss to create the formula. For example, it might use the number of joints and pieces to estimate the setup time for the first truss, and the number of pieces to calculate how long it will take to build the 2nd, 3rd, 4th, etc. trusses – possibly adding a surcharge for long pieces, or 2x6 pieces, or longer trusses. A Time and Motion method always charges more per truss for smaller quantities and less per truss for larger quantities. Despite its obvious strengths, the Time and
Motion method, like its simpler cousins, still involves guessing; the difference is that the guessing involves more variables. More computer generated numbers paired with more Factors. The higher complexity of this method appeals to some, and puts off nearly as many.
As long as we attempt to use one of the “Big Three” to estimate labor (and other price factors) we are playing guessing game. Our Factors may have proven themselves worthy over the long haul, but on an individual job we frankly are going to have very little clue as to the what the factor should be, or why. When it comes to adding things up, computers are king. The “tough nut” is figuring the labor, and a computer examining a truss on a computer screen is innately limited in what it can do. Why? Because it is not looking at several things that we know intuitively are needed to really estimate what the labor cost will be.
What current methods lack
When using standard methods to price, we are not taking into account:
·What trusses will be built together? (what larger production group is the truss part of?)
·How will the pieces be parsed (what saw?) and sorted (in what order?)
·On which work centers will the trusses be built?
Looking at this in more detail, we know, or can figure out, what production groups the job will be subdivided into. We know how we batch jobs for the saw and what our sorting routines are. We know what table and stations we are most likely to build certain trusses on. And we know, or learn over time, the typical metrics for each piece of equipment (setup time per piece, run time per piece (saw) and per truss (table.)
We know all these things, but we don’t use them for estimating. Truss pricing in the future will.
The future
The pricing of the future will be by plant simulation. Once the design work is done (at least the initial pass,) the job passes to the plant simulator which:
·Knows how our saw strategies, creates a batch cutting list on the fly, and estimates the time each part of the process will take
·Knows how we break trusses into production groups, and does it
·Knows what station and table we build each type of truss, knows the metrics (expectancies) for that station, and then applies those expectancies to the work assigned to it
Plant simulation becomes an overwhelming array of numbers unless we can match those estimates up against their real-time counterparts. The “home run” here is to build an estimating simulation that includes only those metrics that you also are measuring in your plant. The more you can (or want to) measure, the more detailed (and potentially accurate) your simulation can be. This is where plant simulation differs from Time and Motion. No guessing; you need to see side by side your estimates next to your actuals.
My example is highly simplified; we would want much more detail in order to understand exactly what the difference was between what we thought would happen and what actually did happen. Plant Simulation is going to become the standard simply because no other method is as practical or accurate.