No liquids in the cabin, no solids in the hold - the Heathrow T5 baggage shutdown

Anyone who follows the news would have heard of the opening day debacle at the new T5 terminal at Heathrow in London England that culminated in passengers only being allowed carry on baggage resulting in a pile of over 20000 bags, wryly described by one blog commenter as ‘No liquids in the cabin, no solids in the hold’. With normality seemingly restored after backlogged baggage sorting moved off to Milan, Memphis and across the UK, what can we say about what happened and could it have been avoided?

Was this ‘business as usual’ for British Airways which is already a leader in the lost baggage stakes (26 bags lost or delayed for every 1000 passengers), having had similar Heathrow incidents (spectacularly in Dec 2006 and as recently as Feb 2008)? While this may have been the popular expectation, the project appeared well planned and executed, with construction of Europe’s largest indoor space (the combined size of 50 football fields over 5 levels) on time and budget, and testing completing weeks and even months before the live date, watched by media witnesses and sundry VIPs. Fifteen thousand pretend passengers with baggage took part in dry runs across a 6 month period, and even though the terminal is more than double the size of the other Heathrow terminals, overnight baggage handling tests got to 90% of expected load. So what went wrong? Not much but the effect was catastrophic. Could it have been prevented? Not sure. Could the risk of it happening been reduced? Definitely.

Most modern terminals open with baggage handling issues, and Robert Atkins, the BA development manager at T5, echoed this, ‘The biggest risk is can the baggage system handle the load’, and if there were problems, ‘We would be in a pickle pretty quickly’. Paul Coby, BA’s CIO, said none of the technology was new but it was a unique implementation, and the project was more of a business transformation so the focus was processes and people. Heathrow is run by BAA and their Heathrow MD, Tony Douglas, said ‘I don’t think there will be major glitches, but once we use the terminal live with big volumes of passengers, we will find out’. The original implementation plan was to move all operations into T5 overnight. This was sensibly changed to moving two thirds, but this still involved moving 1000 objects between 11pm and am, including 95 baggage tugs and 240 cargo containers, in a convoy over 5 kilometers.

Good testing is about managing risk, asking lots of what-if questions. Some of these risk questions can be tested, and some can’t. BA claimed contingency plans ‘to deal with every type of last minute disruption’. That may be true but more focus should have been paid to the first day. There had been 180 separate companies involved in the T5 project, and the development process had run on time and budget. A week after the opening a media campaign would start to publicize the new Heathrow. Staff had probably already seen the ads and were excited about the new era starting on March 28, 2008. Based on this information, what could happen if lots of enthusiastic staff turned up early for work on the first day? There would be more demand for car parking and security screening. What impact could that have? Well we would need to ensure that first shift staff got priority. There Is a carpark for early starters to help them get easy access to the terminal (much closer than the normal staff carpark that is serviced by buses). We need to make sure that only early starters use it. For security screening, ideally we want 2 gates or at least 2 queues at one gate. We can implement these contingencies at little cost and avoid a potentially disruptive bottleneck. Was this considered? Apparently not. Some early starters were forced to park in the normal carpark and there was a single gate that had a queue of 60 people at one stage.

For the first shift baggage handlers not delayed getting in, they discovered they had issues logging in to the baggage handling system. For a showstopper simple implementation error like this to happen to the highest risk system is almost unbelievable. Unfortunately all humans make mistakes but we can take steps to try and avoid them. In this case an implementation checklist of tasks to complete would have avoided the error, perhaps completed by two people working in tandem to reduce the risk even further. The net results of these two incidents was three ‘cut and runs’, planes leaving without any baggage in the hold. Baggage off arriving flights was handled inconsistently. Some passengers on the first flight to arrive (from Hong Kong) had their bags in 5 minutes but some baggage reportedly took an hour. The second flight to arrive (also from Hong Kong) had passengers complaining of bags taking 40 minutes to appear on carousels, but a BA spokesman said ‘It is within the target and obviously we will try to get the bags on the carousel as quickly as possible.’' Shaun Cowlan, BAA Heathrow logistics director, said ‘On the whole it has gone extremely well. Quite clearly there are a few snags but an operation this big is like buying a house. There are always niggles.’

Meanwhile behind the scenes things were not going as smoothly as anticipated. There was confusion with a new handheld system that directed teams to load and unload plane baggage. According to staff reports, teams were apparently being directed to sort and load already canceled planes, and not being directed to unload other planes (passengers reported a team arriving 20 mins after a plane landed in one case. In another, a baggage tug approached a departing plane to turn back with the baggage just before reaching the plane, which finally departed over 3 hours late) Confusion and lack of familiarity with the building had teams arriving late. A new requirement to put baggage in special cargo bins (moved across in the convoy?) also caused delays, with a shortage of bins (did the cancellation process include moving bins to the next scheduled departure gate?) By midday, 20 flights had been cancelled (some with only an hours notice) to try and reduce the baggage backlog. BA had issued a statement acknowledging issues with staff parking and security access, staff familiarisation and ‘baggage performance issues’ (but no mention of the login issues?). At the baggage carousels, some arriving passengers were asking where their baggage was after 90 minutes or more, then 2 or 4 hours as the day proceeded. One passenger reported a 45 minute delay getting off the-plane, then nearly 5 hours to escape Heathrow. BA mobilized desk staff with ‘Can I help you?’ T-shirts and clipboards to walk amongst passengers apologizing and listening to their frustrations.

At this stage any risk analysis would have identified that the afternoon peak was likely to be too much for the baggage system. This may still have not been evident to the baggage supervisors who are located in another part of the terminal and rely on the computer systems for information. Presuming their systems were working, the supervisors were making wrong decisions. They appear to have been working off the original flight schedule not the actual one. That could explain the baggage tug bringing baggage out to load the plane then turning back, if the scheduled departure time had passed and the supervisor wrongly thought the plane had already left. It could also explain some of the other incorrect instructions to the baggage teams.

Could the confusion have been anticipated and avoided? I think so. Bad weather cancellations are common so there must have been some testing of this. They needed to take staff confusion into account. There are 2000 baggage handlers. Some must have been involved in the trials so why not have them work the first shift? Supervisors should have been at hand to closely observe the situation with reserve handlers available to help out, and tiger teams of baggage handlers (perhaps in ‘Can I help you?’ T-shirts?) should have been in place as guides and trainers to assist where required. At 5pm, the baggage system overloaded and shut down. Shortly afterward Gareth Kirkwood, BA’s director of operations, announced the suspension of baggage check-in, saying ‘We always knew the first day would represent a unique challenge because of the size and the complexity of the move into terminal 5.’ In total 68 flights were cancelled on the first day, 72 on the second with similar numbers since then. An apparent media lockout started on the second day, in contrast to the media blitz on day one.

Finally on the 12th day (after nearly 500 cancellations and 11000 SMS text message apologies [privacy laws only allowed passengers who had consented to messages to get them!]) a full scheduled day without cancellations occurred, though they had announced this for the 8th day only to be frustrated by a baggage system bug and bad weather. These numbers seem to indicate that while 60% of operations moved into T5, it was only able to function at 50% capacity initially. Surely this was modeled before the move and understood, so why did they go with 60%? The plan to bring the remaining flights over at the end of April is now moved to June. The company responsible for the baggage system machinery has said it was working properly but there were issues with its operation. BAA stated, ‘The baggage system did not fail. The system is designed so that when it is full, it stops so the backlog can be cleared. Bags were on the system and were not being taken off’, but later statements by BA seemed to regard BAA as a significant part of the problem (which my analysis seems to support). A fortnight after the opening, the BA Chairman, Martin Broughton, finally commented describing the incident as ‘An acute embarrassment’ and explaining ‘It is very difficult to itemize what went wrong- it is a complicated set of things rather than one thing, but you have to focus on the baggage system as the key thing that has not worked according to plan. Many of the other things are typical teething problems’. In early May, the British Government will hold a formal enquiry.

A guidebook to T5 begins ‘27th march 2008. An amazing moment in history’. Graffiti in the men’s toilets simply says ‘Welcome to hell!’. Much of the media and blogosphere has been trumpeting project mismanagement and poor testing. As I have argued, I think the main cause was poor risk management of the first day of operations. One blog commentator said ‘Shades of Denver’, a baggage system that was decommissioned after 10 troublesome years. That is an unfair comparison. After 3 days, all but 6% of passengers had their bags and none were torn apart like Denver! As T5 opened, Veronica Kumar, BAA’s Head of People and Change, said ‘Our policy has been to create a context for change, then to apply changes within that context. We want to give flyers an experience they’ll remember’. It will be remembered, and for some never forgotten! Better planning for day 1 could have completely changed that experience. At this point it is unclear exactly when the media advertising campaign on the speed and efficiency of T5 will be broadcast. (grin)

In June, just as T5 (hopefully) gets to full capacity, a new book Heathrow’s Terminal 5 – History in the Making will (hopefully) be released. It describes “successful delivery of this mega project on time, on budget, and safely”. The Operational Readiness section is online and it describes the thoroughness of the project. There was a test center that all software had to go through from early 2004, and a typical airport slice with one checkin, one gate, one stand etc from early 2007. While I have taken a light approach to the fiasco, it is tragic that poor implementation has destroyed all the good work. Perhaps they could use this blog post as a postscript to the book?

At the end of the day it is all about getting baggage on and off planes. When the baggage system works correctly, it is very fast and efficient. Don't believe me? Hang on to your seat, and watch this bag’s eye view video.

(Sources – dozens of articles including most of the UK daily press (The Times, Daily Mail, Telegraph, Independent, and Guardian [fantastic reporting of day 1 issues here ), Computerworld , Computer weekly, Economist, Bloomberg, Silicon, BBC, Sky News, MSNBC, L A Times and the blogosphere.)