In the final stage of our ranking list, we reach number three that corresponds to the most important designer of the last 30 years. This is none other than the American Burt Rutan (1943) who from his company Scaled Composites in Mojave’s dessert, with his revolutionary developments and with the use of composed materials, has been able to proof that the innovation in the aeronautic world is not only in the hands of companies like Boeing or Airbus, but in the hands of people with an extraordinary talent and his small groups of engineers.
Among his achievements we find the design of the GlobalFlyer that made a non stop flight around the world with a single pilot, the Ares, a revolutionary low-cost jetfighter, the StarShip, designed for Beechcraft as an executive jetliner in canard configuration, that employed for the first time a carbon fiber composite for the structure and, of course, the design of the SpaceShip One (and the carrier aircraft White Knight) which won the X-Prize in 2004 flying at an altitude of over 328000 ft.
Today we will talk about one of the most difficult duties that one has to face in his professional life: write a good specification. Although, as mentioned previously, this work is usually given to the newcomers, this task is one of the most complex ones that an engineer can face. It takes knowledge, experience and synthesis capacity and these are capabilities that not everybody has.
When someone faces the problem of developing a specification of an equipment or system, the first step is the deep understanding of the unit’s utility. Once accomplished this, the challenge is to explain this to everyone else.
What we are going to discuss here is not related to any of the several software tools that help with the specification’s management. We will see something much more basic. How these requirements are related with the common sense.
Unfortunately for us, most of the specifications are nothing more than a huge pile of unstructured requirements, confusing and often contradictory or impossible to be reached. So if you ever have to write a specification, think about how much you suffered with those of other people and try to apply what you learned in order to avoid the same mistakes.
A good specification has not to be as thick as a brick. I have seen excellent complex system specifications and the requirements were collected in 20 pages.
The first item that a specification must have is a general description of why is the equipment needed (functionality) and under which conditions it has to work (environment) and its relationship with the outside world (context). This will allow the designer to have an overview of the problem from the start, which is a big step.
Most of the specifications we find, do not fulfill what we mentioned before and after two pages full of unpronounceable acronyms, listing a punch of infinite requirements in no particular order, you find yourself obliged to jump from page to page and half an hour later you don’t know if you should design a power source or a HF communication device.
The first things that attract you attention are the obvious ones. It is common that the first requirement of a specification of any equipment contains a sentence like:
The equipment shall operate normally when used in its operational environment
To a sentence like this we could add the following ending:
Otherwise it could be used as paperweight
Let’s see. From a formal point of view, this may have a deeper meaning, but it is obvious and wrong that the specifications are full of them. But beware, there is no single harmless requirement, as sometimes, an apparently futile requirement can have important design problems.
Most of the specifications are like Santa’s wish lists, as children are setting objectives that afterwards turn out to be unachievable, but afterwards, “a posteriori”, are modified to adapt them to real life, in order to keep harmony.
It is common to find requirements that say something like this
The maximum power consumption shall not exceed 250W
And you, as designer, look surprised and ask yourself, how is it possible that the one who wrote the specification could know which could be the energy consumption of an equipment that has not even been designed yet.
For this question there are only two possible explanations:
- First, the person who wrote the specification knows very well what he is talking about and which is the energetic availability for this specific equipment that cannot be exceeded, and if this target cannot be accomplished the whole project would have to be abandoned.
- Second (the most common), the “specificator” makes a pie in the sky and sets a random value, usually taking the example of a similar equipment, taking special care of add the letters TBC (To Be Confirmed) at the end of the requisite. Then, if the final consumption is 280 W, the specification is modified to 300W and the tag TBC is deleted. So everything fits in.
Other times the problem is to specify more than necessary, leading to paradoxical situations. For example a specification could require that your equipment communicates with another unit and a point to point communication bus will be necessary. In another paragraph it is said that the cycle frequency should be 100 Hz and that the size of the data is 100 bytes.
Up to this point everything all right. But then you start doubting, when four pages later, it is specified that the transmission clock shall be of 15 MHz. You ask yourself, why the hell do I need a 15 MHz clock (which will give me a lot of trouble at processor and EMI level) when I could perfectly work at 150 KHz and I still have free time?
At other times you are simply astonished when you read something like:
The equipment will be resistant to acetone
And then you imagine that this requirement is set in case the pilot is a woman, who likes to be always pretty and when the autopilot is on, she renews her nail polish and then the acetone bottle may spill over your equipment.
When you face a specification you should make it with sportsmanship and imagine that you are running the 400m hurdles, followed by a half marathon. You should not discourage even if you stumble and most important, you should not get exhausted.
A thought:
An expert is a person who knows more and more every day about fewer things until he knows everything about nothing.
See you
No comments:
Post a Comment