Home > Writings > Blog Overview > My Blog

Complexity

29 April 2011 | Technology

Today's planned launch of Space Shuttle Endeavour was cancelled earlier because of a problem with a fuel line thermostat for one of the auxiliary power units that drive one of the hydraulic systems. This delays the one but final Space Shuttle mission for a few days. The very last Shuttle flight will be when Atlantis completes the Shuttle programme with a mission in June or July. I remember seeing the first launch of a Space Shuttle into space, back in '81, as well as the excitement I felt at the time. I do feel rather sad that the Shuttle programme comes to an end. The Shuttle has played a key role in our continued attempts to explore, in Carl Sagan's words, the shores of the cosmic ocean.

While now, as the programme is coming to its conclusion, public attention for the Shuttle is once again booming (as was the case in its early days) for most of its existence the interest of the general public in space exploration has been rather muted. The progress of science and technology always seems to find it hard to compete with celebrity gossip, X factors or other such twaddle. Despite two dreadful accidents, we have seen considerable progress in the past few decades and the Space Shuttle can look back on a highly respectable track record.

At the same time, there is an interesting lesson in the Shuttle programme's fate, underlined by the accidents that destroyed two of its vessels and their crews. Despite the ambitions to create a reusable space exploration vessel, the intricacy of the Shuttle design made it expensive and complicated to operate. The 1986 Challenger mission ended in disaster when the orbiter broke apart 73 seconds after its launch due to the failure of an O-ring on its right solid-fuel rocket booster.

The lesson is that complexity comes at great cost and thus should be avoided where possible. Apparently, the reason for designing the solid-fuel boosters in segments, necessitating sealing rings, was in major part due to transport logistics. If the booster rockets would have been designed as a single tube, Challenger would not have been lost.

The added complexity of the booster design with O-rings was therefore a key factor in the 1986 disaster. Similarly, many problems I come across in my daily practice can be traced back to unnecessary complexity of IT solutions. That is why I stress so much the need to keep designs and implementations simple, easy to understand, to manage and to maintain. Space exploration is pushing the boundaries of our scientific and technological knowledge. It often comes with great complexity. In contrast, addressing many real-world business problems through well designed, clear, simple solutions is often not rocket science. Instead, it is a matter of common sense and clarity of thought.

Letís value intelligibility, simplicity and transparency a bit more. If done right, one will notice that simplicity does not result in less sophisticated or user-friendly systems. But it does result in more robust, more usable and more sustainable solutions.