The article continues so please keep-up reading. There will be much fun and
deep philosophies later on text...

"2. DO think "cost."  Someone said that the difference between science,
where discovery is so important, and engineering, where a product is of
prime importance, is cost. (Cost is important in science, too, but it's a
secondary consideration.) So try to estimate the cost of each project, each
design, and each deliverable item.

Start out by finding how much you cost: your salary and the "fringes." By
the way, the fringes-holidays, vacation, health insurance, education,
retirement, and so on frequently amount to another 30-40 percent of your
salary. Then add in other overhead costs like light, heat, and real estate
taxes plus the G&A (the general and administrative costs for "those guys in
corporate") and even some profit, we hope.

DO find out how much the things you use cost. What did your organization pay
for that sampling oscilloscope? How much for the new workstation? How much
for that software package? Keep a notebook and create your own cost
database. Get acquainted with your employer's cost database. (They sure
better have one!)

There are three ways of making an early cost estimate for your project:

   - The bottom-up approach, based upon a detailed estimate of the
   design, fabrication, and materials costs of each item in the system.
   - The comparison approach, based upon the known costs of similar
   previous efforts. For example, the new system consists of a transmitter that
   has about twice the complexity of the one used on project A, an antenna
   about one-and-a-half times the challenge of the one on project B, a software
   package with about half (watch out!) the code on project C, and so on.
   - The rule-of-thumb approach, based upon experience over a number of
   jobs and companies on cost per pound, per printed circuit card, per line of
   code, and so on.

The comparison approach supplemented by some rule-of-thumb checks is usually
the most accurate and takes the least time for an estimate early on. I've
never seen anythmg yet that couldn't be broken down so as to relate a large
part of it to projects that had already been done. That even goes for
projects you believe you'll be starting from scratch. It can even apply to
integration of an entirely new system. So DO get familiar with cost results
on prior jobs
and start creating some rules of thumb of your own."
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist