Jakob Østergaard Hegelund

Tech stuff of all kinds

Failure

2012-06-06

Much of the technology we surround ourselves with every day in our lives is built as relatively complex constructs made up of smaller simpler parts. A car for example - a hugely complicated machine with thousands of relatively simple parts (like a tyre for example - most non-technical people will understand the basic concept of what the tyre does without understanding the full construct of the car). Or a house - most people understand that the roof keeps the rain away from the mattresses while few may understand how cavities and reflective surfaces influence the energy requirements for heating the house during winter.

That is all well and good - these everyday products function alike. Cars of various makes and models all perform the same basic function even though they may look, sound, feel and perform very differently at wildly different price points. The same holds true for the other example, houses - they all generally keep the rain from the mattresses while other factors can vary.

In other words, there is a set of basic functions that one can expect a given everyday product to perform, even if one does not understand the detailed making or function of the product. For example, I can go and buy a leather jacket and expect the leather to be soft and durable, even though I have no idea how leather is cured and how a jacket is made so that it is both comfortable and durable. The same will hold true for a car, a house, a blender, a coffe grinder, a kitchen knife, a bicycle or a pair of shoes. Or any other everyday product. We come to expect some basic functions to be delivered by everyday products and this is what makes our modern way of life possible. It makes it possible for me to buy a pair of shoes without being a skilled shoemaker, and still expect them to keep my feet dry.

This brings me to the topic of failure. When an organization manufactures and delivers a product that completely fails to perform the basic function that the consumer expects of the product, then it matters. It matters because such a failure is not isolated to one level of the organization. As discussed earlier, everyone understands the basic requirements of everyday products without necessarily understanding the detailed technical constructs that make them possible. Therefore, there is no excuse for a non-technical sales person to sell a product that cannot perform the basic function it is meant to perform. Being non-technical is not an excuse. The door swings both ways; the technical person cannot excuse him/herself with "just delivering what was sold". If you know that the basic functionality is not met, then you know that nobody could ever want that product. Not like that. Not without basic functionality.

We recently moved offices to a new location which is generally really nice - with one important exception... The elevators. There are two elevators for eight floors of offices and there is very poor access to stairs. One would therefore expect that there would be a good utilization of the two elevators.

Let me just take a step back here. You have probably used elevators before. You know how it works - you press the up arrow if you are going up, or you press the down arrow if you are going down. There are a number of elevators next to one another, and pressing a down arrow will make the "next available" elevator that is passing your floor in the right direction (downwards) stop to pick you up. This is simple and efficient, it works extremely well, and this is what you would exect elevators to do. It is, in other words, basic functionality when you have more than one elevator.

Well back to the topic; the two elevators we have are not interconnected like that. So, when you press the "up" arrow in the morning, only the elevator for which you pressed the "up" arrow, will consider picking you up. You have no way of knowing which of the two elevators may be moving in which direction any time soon (you can see their current floor but that is it). So, the elegant solution that everyone employs whenever they need an elevator, is, to press the buttons on both elevators.

While this may seem to solve the problem - it does request both elevators individually to pick you up - it does not take a rocket scientist to compute that more often than not one of those two trips that you order will be wasted as you obviously can only use one elevator at a time even though you request two.

Effectively, we have slightly more than one elevator - 1.1 elevator maybe. Not 2 elevators by any stretch of imagination. There is maintenance of two elevators, energy consumption of two elevators, space utilization of two elevators, but only effective capacity of slightly more than one elevator - because almost everyone always requests both elevators to pick them up.

This brings us back to the topic of failure. Everyone, from the guy installing the product all the way back to the sales team who sold the solution - they have all failed completely. The elevator "computer" obviously supports cooperating - I have used many elevators in many countries and I have never experienced non-cooperating elevators anywhere. Most likely the "cost" of making this cooperation work is a few bucks worth of copper wire to interconnect the two elevator computers. The cost of not having it work, is the complete cost of operating an elevator that is making trips nearly non-stop all day while not effectively transporting anyone anywhere.

The guy who tested the elevators knew this. The guy who installed the elevators knew this. The guy who sold the elevators knew this. The guy who managed the contracts knew this. Yet, everyone accepted that this everyday product would completely be failing to deliver the basic service that everyone knows that everybody would expect.

Aside from just being a rant, I actually wanted to point out how this is a sign of a failing product delivery organization. When blatant failure can be ignored (I refuse to belive that it has not been noticed) at so many levels of the organization, something is simply wrong. It might be interesting to observe how this particular elevator company will be performing relative to its competitors over the next decade. It is indeed possible that the failing organization is local to the capital area for example, so it may not have devastating impact on the company as a whole - I still think it is interesting to reflect on the level of willful ignorance necessary for a product delivery to fail so remarkably on basic expected functionality.