Interchangeable parts - Why not software?

The concept of interchangeable parts is one that packaging machinery builders readily understand and depend upon every day. Not only are parts from a single manufacturer interchangable within a machine, but builders source custom parts from multiple manufacturers to control cost and schedule and generic parts are interchangeable across a host of different machines and devices. We have come to enjoy and expect that fasteners, shafts, gears, couplings, motors, valves, and many other generic parts are fully interchangeable. This evolution of the concept of interchangeable parts, introduced over 2 centuries ago, came about with the adoption of proprietary, regional and international standards. Interchangeability leveraged advances in technology and brought opportunities and competitive advantage for manufacturers.

With the concept of interchangeable parts providing such great benefits to packaging machinery builders and users alike, why then do we shun concepts and standards that could provide the same benefits for software? Why aren't we demanding more interchangeability of software?

The online encyclopedia WikipediA says that: "Interchangeable parts are components of any device designed to specifications which insure that they will fit within any device of the same type. This streamlines the manufacturing process, since all pieces are guaranteed to fit with all others, and it similarly creates the opportunity for replacement parts."

A future WikipediA entry might say something like this: Interchangeable PackSoft objects are software components of any packaging machine designed to specifications which insure that they will operate within any machine of the same type, regardless of the controller. This streamlines the software development process, since all the objects are guaranteed to fit with all others, and it similarly creates the opportunity for replacement parts, easy upgrades and functional enhancements.

We can always learn from history and apply that learning to today. Continuing with WikipediA's description of the evolution of interchangeable parts, I invite you to read the following paragraphs from WikipediA, substituting my words in [ ] for WikipediA's words in ( ).

Prior to the (18th) [21st] century, devices such as (guns) [packaging machines] were (made) [programmed] one at a time by (gunsmiths) [programmers], and each (gun) [machine] was unique. If one single (component) [program] of a (weapon) [machine] needed a replacement, the entire (weapon) [machine program] either had to be sent back to an expert (gunsmith) [programmer] to make custom repairs or discarded and replaced by another (weapon) [program].

Around (1778, Henri Blanc) [1992, PLCopen members] began producing some of the first (firearms) (machines) with (interchangeable parts) [interchangeable software objects]. (Blanc) [The PackML organization] demonstrated in front of (a committee of scientists) [attendees of PackExpo] that his (muskets) [packaging machines] could be assembled from a (pile of parts) [library of software objects from different vendors] selected at random.

In the U.S., (Eli Whitney) [the OMAC Packaging Workgroup] saw the potential benefit of developing ("interchangeable parts") [specific software objects] for the (firearms of the United States military) [packaging machines of the United States packaging industry], and thus, around (1798) [2002], he built ten (guns) [machines], all containing the same (exact parts and mechanisms) [PackTags, state model and PackSoft objects], and disassembled them before the (United States Congress) [PackExpo attendees] . He placed the parts in a large mixed pile and, with help, reassembled all of the (weapons) [machine software] right in front of (Congress) [the attendees].

The (Congress) [organization] was immensely impressed and ordered a standard for all United States equipment. With (interchangeable parts) [standardized software objects], the problems that had plagued the era of unique (weapons and) [packaging] equipment passed, and if one (mechanism) [module] in a (weapon) [machine] failed, a new piece could be ordered and the (weapon) [machine] would not have to be discarded.

The principle of (interchangeable parts) [standardized software objects] quickly made mass production in all sorts of industries relatively easy. It was based on the use of templates, applied by semi-skilled labor using (machine tools) [programming libraries] instead of the traditional (hand tools) [custom software]. With the introduction of the (assembly line) [object modules] at the beginning of the (20th) [21st] century, (interchangeable parts) [standardized software objects] became ubiquitous elements of manufacturing.

Will we learn from the history of interchangeable parts? My guess is that colleagues told Mr. Blanc that he was a dreamer- it was impossible. What do you say?

TrackBack

Comments Great article Keith. I got turned on to the idea in the 1980's when Brad Cox spoke about software interchangeability and the potential to mimmic what we see in the industrial revolution (Henry Kaiser Ships, Eli Whitney etc) A great challenge indeed, but via Object Oriented techniques, we can build the Geodesic Dome of software application development, which would yield this assembly-line efficiency in software development, same as modular construction of skyscrapers today. Once you start the intelligent assembly of interchangeable parts, you can't go back to traditional software development! Posted by: Jesse Tayler on March 30, 2008 Post a comment (If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.) Name: Email Address: URL: Remember personal info? Comments: (you may use HTML tags for style)