I have been trying to follow the problems in the new T5 terminal at Heathrow Airport. There are small project management problems and there are big ones. Ours are modest. On the scale of big, you have T5 which apparently included 400,000 hours of software development (that’s a lot of lines of code!), a full year of testing and a full year of training. Yet its open was a technology disaster.

From the various press articles I have read, I can only imagine what happened. There were so many vendors, so many new hardware devices, as well as new conceptual frameworks for storing and manipulating data.

Here was one of the bits of bravado from Nick Gaines, the IT director for the British Airports Authority, forecasting a good launch. "Integration is not just plugging technology together; it’s about people, processes and systems working together." Tough talk from Nick.

Apparently, British Airways underestimated everything from training, to testing, to phasing in the new systems involved in the T5’s unprecedented level of computer integration of baggage handling. And they had planned that coding was incomplete and would continue even after the terminal opened.

Aside from the baggage mess, the other big problem had to do with plans to use temporary fingerprints as a means of identification. There was such a reaction to this, they had to back off. One has to ask, how could they have not anticipated this reaction and field tested it with real population groups in advance? If they couldn’t anticipate there might be social and political unease with universal biometic fingerprint gathering, it is not hard to imagine they also didn’t get the baggage and other systems right either.

The T5 experience illustrates that we have fully moved into an era where most large software projects are not one system built from scratch (T5 has 180 suppliers, 163 systems, etc). We are in an era of multiple systems combined together with complex data exchange and integration. More things, more different technologies, and more people have to work together in more complex ways. Problems will happen.

We’re thinking a lot about project management ourselves these days. I guess the lesson of the T5 open is that no matter how much planning time you put into an IT project, you may need more. This is certainly hard to explain to clients. Scott Berkun, author of my favorite book on this topic, "The Art of Project Management" from O’Reilly, says something to the effect that no matter how good your systems are and how much time you devote to them in a software project, the main measure is how effectively you can respond to the project crises which arise no matter what. Sobering. Ask us about what we’re doing about it these days.