System update kicks Heathrow baggage system offline
The computerized baggage handling system in Heathrow Airport's Terminal 4 was kicked offline for about two days due to software problems. As a result, carriers in the terminal were forced to sort baggage manually and 6000 passengers experienced delays, cancellations, and hassle.
A terse press release from BAA, which operates Heathrow, stated:
The performance of the Terminal 4 baggage system has...been reduced by an ongoing software problem....[P]assengers should consider minimising the amount of hold luggage they check in.
BAA added:
"We have two sorting machines, one is a redundancy, and we have never had both go down before, that is why it was such a significant problem for us. We are still analysing exactly what caused this at the moment."
The error occurred, "[A]fter an update to the system made it reboot...." Additionally:
BAA told Computerworld UK that a computer software upgrade, conducted Monday night, was the cause of the failure. Despite testing the software, the "glitch crashed both baggage sorting machines...."
THE PROJECT FAILURES ANALYSIS
IT plays an integral, if often hidden, role in the transportation landscape. Transportation systems often contain standard IT components -- clients, servers, hardware, software, networks, and so on -- and are therefore subject to the kinds of flaws and breakdowns faced by corporate IT systems.
For those interested in the IT complexity present in baggage handling systems, see this system diagram from Logan Teleflex, a leading supplier of such equipment:
In addition to heavy-duty IT, modern bag-handling systems are complex, highly-mechanized, computer-driven operations involving many physical components and subsystems. Given all the moving parts, I'm sometimes amazed these systems work at all.
In the Heathrow case, it seems both machines in the redundant system were upgraded at the same time. Since the two machines failed to restart properly, evidence points to a poorly-tested upgrade as the cause of failure, despite BAA's comment to the contrary.
Two questions arise:- Why wasn't the upgrade tested more fully before deployment?
- Why didn't the upgrade technicians ensure the enhancement worked on the first system before applying it to the second?
There are no good answers, which makes the software supplier look pretty darn bad.