Editor’s note: Eric Jackson, the founder of DeepWeave Technology Consulting, has built his career pioneering software solutions to particularly large and difficult problems. He will be writing a regular column for LTW titled “Going Deep.” It will be published on Tuesdays.

BLACK MOUNTAIN,Analogies are always a dangerous predictive tool, but they are also often useful, particularly if their predictions are backed up by other considerations.

I have been thinking lately about the parallels between the software industry and manufacturing, on a number of fronts. One parallel is currently very much in the news, namely the accelerating movement of software development jobs offshore, following the pattern set by manufacturing. I will be taking up this topic, together with its implications for education and workforce development, in a future issue.

What I am struck by at the moment, however, are some parallels in the historical development of manufacturing and software engineering. Although the production of goods has much older roots, both software engineering and the manufacture of goods came originally out of a culture of craftspeople, one in which each item is individually made by a skilled artisan. Over time a storehouse of knowledge was accumulated and shared, generating strong traditions of technique and craft wisdom.

Software without wizards

With the advent of industrial age in the 19th century, manufacturing became increasingly concerned with the development of precision interchangeable parts, standards, and processes to support mass production of high-quality products. We find this very much echoed in software engineering in the increasing concern for standards, code reuseability, software engineering processes, etc. To some extent, the movement offshore of some of these jobs is a measure of the success of these efforts.

Software is no longer the purview of strange geniuses and wizards – it is increasingly a task of design followed by implementation by technicians who follow the blueprint. It may well be that the future will also see computer-generated programs, paralleling the automation of factories in the manufacturing arena.

There is, however, another interesting opportunity that presents itself from the analogy with manufacturing. A recent and interesting theme in manufactured products is that of mass customization. It is a theme, rather than a technique, since it takes wildly different forms. Examples range from Dell’s customizable web-based model (very simple, but requiring sophisticated logistics to back it up) to the custom manufacturing and inspection technologies of the NC start-up Raindrop Geomagic (which uses very sophisticated 3D software).

Profits and personalization

What is common is the realization that significant value can be created by returning to customized or personalized products if the economies of mass production are retained at the same time.

I believe a similar opportunity exists in software, both personal and business. The days of profitable shrink-wrapped personal software are undoubtedly gone, but what about really personalized software – data storage that molds its organization to your way of looking at things, search aids that learn your preferences and use them to tailor your view of the digital universe. In business, I wonder how much potential productivity gain is lost through the inflexibility of current software models, which require the business to mold itself to the software more than the other way around.

What will it look like? I have no idea. As in the manufacturing arena, what is important is the idea itself – there is undoubtedly a huge variety of specific approaches. Some possibilities include systems that learn and adapt as they are used, layered and modularized systems that allow flexible connection of modules and insertion of new layers, parametrized systems like many that exist today, but with a more sophisticated model allowing rapid (and automatic) customization based on installation information. Some will be intimately tied to a specific product, others may suggest more general uses.

I would suggest, however, that in the spirit of the general acceleration of everything in the world, we won’t be waiting another 100 years to find out.

Eric Jackson is the founder of DeepWeave. He has built his career pioneering software solutions to particularly large and difficult problems.

He began by designing computer codes to solve turbulent flow problems using the most advanced supercomputers and parallel supercomputers available. After 10 years on staff in the Applied Math department at Princeton University, he moved to a private consulting firm in Princeton where he designed and implemented one of the first and most sophisticated software applications to automatically analyze and correct computer chip designs to improve their manufacturability.

In 2000, Eric co-founded Ibrix, Inc. He is the inventor of the Ibrix distributed file system, a parallel file storage system able to scale in size and performance to millions of terabytes.

DeepWeave: www.deepweave.com