I once had a customer pursuing a “requirement” to process a million rows of data in a set period of time. This was deemed a critical component of a user acceptance test for a new system. By itself this was not an issue, but the question in my mind was, “why would anyone need to bring back a million rows of anything”.
The inference was that the user needed all that information to answer a question and the additional implication was that somewhere in the million rows of information was the answer to some question that the user really needed to find. So the reality was that the user’s question was not “I need a million rows of information”, but instead was “I need to sift through a lot of information to find the answer that I seek.” My instinct told me that perhaps a more specific question was appropriate in order to save the user some time and get to the answer more quickly.
A related requirement was to bring back all queries in less than a certain period of time. This did not apply to just tactical transactions (that really needed near-real-time responses), but all system queries including strategic analytical data sets. How, then, was a system requirement developed to bring back so much data or deliver an instantaneous answer to everything. Something was broken in the requirements process.
The PBS series “Connections” was always interesting for me, because it joined the dots between present-day wonders and the science and invention required to make them so. It showed how various discoveries, scientific achievements and world events were successively interconnected to bring about particular aspects of modern technology. For a requirements developer and strategic planner, it provided interesting perspective.
Do you know where the requirement for the height and width of the USAF’s C-17 was derived? Among many other large items, the C-17 had to carry an Abrams tank, so it needed to be so many feet high and wide.
That is interesting bar conversation, until you ask where the Army derived the height and width of an Abrams tank? It could have been wider and higher, except for the one very important requirement that it be able to maneuver the terrain in Europe, which included passage through most train tunnels. So a major planning component of the Army’s battle force was driven by the need to leverage the train landscape.
But, a question then can be posed, what drove the height and width of the train tunnels? The obvious answer is the height and width of the trains, but what drove that? The men who created trains didn’t look too far when they wanted to leverage the steam engine to power their new transportation vehicle The wagons of the day would do nicely, if they were modified a bit, so these men simply put the engine on a wagon, put a stack on it to keep the smoke out of their eyes and the world of train travel was begun.
In most parts of the world travel by wagon involved creating roads that following the terrain, over the plains, rivers and mountains. Sometimes the grade on the mountains was too difficult to traverse, so was begun the backbreaking work of going through the mountains rather than over them, and that involved digging out tunnels that would fit the bill. So the requirement for the road tunnel wide enough for a wagon to pass was established
Again interesting bar conversation, but one might ask where the requirements for most of the wagons of the day was established. Wagons were made of mostly wood and some important metal pieces (such as the axle), and there needed to be some consistency between wagons and carriages as people began to travel about Europe. They needed to know that wherever they traveled they could repair their vehicles without having to carry a whole new vehicle with them. Over time they adopted some standards for wheels and axles that made sense and used what was in place.
Interestingly, hundreds of years prior to the world of trains, people were using the interstates of Europe that were put in place by the engineers and craftsmen of the Roman empire. The Roman engineers establish standards that enabled the field troops to quickly build a roadway network throughout Europe as the Roman Legion needed to quickly move and put troops in place anywhere within the empire. Their solution was to engineer roadways to carry the primary vehicle of the legion, which was the Roman chariot. So the roadways were built to accommodate the width of a chariot.
But what determined the width of a chariot. This was an entirely tactical requirement — a driver and an archer needed to be pulled by two horses for greatest efficiency on the battlefield. And since two horses could not be made smaller than God made two horses, the width of a chariot was derived from how closely those horses needed to be bridled together. Or from a somewhat different perspective, the width of a horse’s ass drove the width of the chariot. So for practical purposes, the width of a C-17 was ultimately driven by the width of a horse’s ass.
Sometimes, from a defense acquisition perspective, there is a good reason to ask operations folks where the requirements for a new weapon system are derived. There may be a long, but valid reason. On the other hand, there may be a more direct way to get to the specific need that saves time and resources and more efficiently answers the mail. Technology enables us to do and think about things differently than we could have imagined only a few years ago, and often allows us to go after specific objectives directly when the same was not possible in an earlier time.
In World War II, during two raids against the ball bearing factories in Schweinfurt, it took 521 B-17s, carrying well over 2.5M lbs of bombs and with aircrews of over 5,000 airmen to fly the missions with at least an 80% certainty of effectively destroying the target objective. During the First Gulf War, technology enabled one F-117 with one pilot and one missile to complete the mission with a 95+% probability of destroying the target. Today, it doesn’t even require a man-in-the-cockpit.
Moore’s law observes that the number of transistors in a dense integrated circuit doubles approximately every two years. Dennard scaling suggests that power requirements are roughly proportional to the area for transistors, and the combined result is that as circuits get faster (i.e., more transistors in a circuit), their power consumption stays about the same. As circuit performance increases, so too has the ability to put more data points on storage media, and hard disk and other storage cost decreases. Network capacity increases and the cost of data transmission decreases. What does all this infer? The bottomline, for those of us who require a less technical explanation of these phenomena, is that technology (especially information technology) changes significantly every 18-24 months enabling new ways to do things that were previously not possible.
We use information and data to make decisions do business in better ways. The issue for those in the technology business is in being enamored of the technology and forgetting about the business or mission requirement. Can we bring back a million rows of anything or provide an instantaneous answer to everything? Probably, and if not today, perhaps tomorrow or in the not-too-distant future of technology. But what is the cost to make that happen, and is it worth paying that price today if the result is not really what we need to achieve to improve the business.
Defense acquisition is a complicated give and take of requirements, priorities, funding, technical capabilities, and mission objectives. It is an amalgam of cost, schedule, and performance, all stuffed into a risk management balloon. If you squeeze one part of the balloon, it will respond and expand outward in another area. Technology advances suggest that we can ultimately accomplish objectives that may now be impossible, but at some cost. Understanding the requirement upfront, including where it was derived, is critical to understanding the risk in getting there in a reasonable manner.
Don’t be afraid to ask the question.
shapethevision 