The History Of Baggage Handling Systems

What was to be the worlds largest automated airport baggage handling system, became a classic story in how technology projects can go wrong.

Faced with the need for greater airport capacity, the city of Denver elected to construct a new state of the art airport that would cement Denver’s position as an air transportation hub.

Denver’s new international airport was to be the pride of the Rockies, a wonder of modern engineering [1] . It was almost twice the size of Manhattan and about 10 times the breadth of Heathrow. The airport can handle the landing of three jet planes simultaneously even in bad weather. The airport’s baggage-handling system was even more impressive than its girth. The coal-mine like cars ran along 21 miles of steel track. 4,000 “tele-cars” routed and delivered luggage between the counters, gates and claim areas of 20 different airlines [2] . A CNS of some 100 computers connected to one another and to 5,000 surveillance cameras, 400 wireless receivers and 56 bar-code scanners coordinated the safe and timely arrival of every baggage.

At least that was the plan.

The airport’s baggage handling system was a critical component in the plan. By automating the baggage handling, aircraft turnaround time was to be reduced to as little as 30 minutes. Faster turnaround meant more efficient operations and was a cornerstone of the airport’s competitive advantage.

Despite such noble intentions the plan promptly thawed as inefficient estimation of the complexity of the project resulted in swelling problems and public mortification for everyone involved. The inauguration of the airport got delayed by 16 months because of the problem. Expenditure to maintain the empty airport and interest charges on construction loans cost the city of Denver $1.1M per day throughout the delay.

Of the numerous awkward gaffes along the way one was an unplanned demonstration of the whole system to the media. It elucidated how the system crushed bags, expelled content and how two carts moving at high speed retorted when they bumped into each other. When opening day finally arrived, the system was just a silhouette of the actual plan. Rather than computerizing all 3 concourses into one stand-alone system, it was used only in one concourse, by a one airline and only for flights which were outbound. Rest of the baggage handling was implemented using standard conveyor belts and a tug and trolley system that was entirely manually operated. This was hurriedly built when it became absolutely clear that the automated baggage handling system was never going to achieve its purpose.

Although the offcuts of the system survived for 10 years, the system never worked properly and in August 2005, United Airlines announced that they would forsake the system completely. The maintenance costs of $1 million per month considerably exceeded the monthly cost of a manual tug and trolley system.

BASIC MODE OF FAILURE

Like all other failures the problems can be observed from numerous vantage points.

In its humblest form, the Denver International Airport (DIA) project botched because the key decision makers could not analyze the complexity of the project with perfection. As intended, the system was one of the most complex baggage system ever endeavored.

There was an exponential growth in the complexity of the project as it was almost ten times larger than any other baggage system. The heart of the complexity was a problem related to what is known as “line balancing” in project management terms. To adjust system performance, empty carts had to be circulated around the airport ready to carry fresh baggage.

With more than 100 pickup points, the algorithms necessary to anticipate where the empty carts should wait for new bags epitomized a hideous dream in the mathematical aspect queuing theory.

This failure in anticipating the number of empty carts required led to a delay in the picking up of baggage an as a result of which the system performance slide downhill.

One of the main reasons of delayed initiation of the project was failure to recognize the complexity and the risk involved. Based on the original project schedule, this delay left a little over two years for the service agreements to be signed and for the system to be designed, developed, tested and implemented. The closest analogous projects although much smaller and simpler, took two years to implement. Given the dramatic increase in terms of size and complexity, implementation the Denver system in two years was an unmanageable task.

As a result of the erroneous estimation of the complexity of the baggage system, the efforts required were also underestimated. That meant that without comprehending it, the Project Management team had calculated the baggage system as the critical path of the whole airport project. To meet the airport’s scheduled opening date, the project needed to be accomplished in just two years. This obviously was inadequate time and that miscalculation resulted in the project being exposed to gigantic levels of schedule stress. The other succeeding problems were most likely a result of (or aggravated by) shortcuts opted by the team and the mistakes committed by them as they desperately tried to meet the schedule.

KEY DECISIONS THAT LED TO DISASTER

Till now what has been discussed has made the basic mode of failure pretty clear. But to reach the heart of the problem and what could have been done to differently, we need to understand how the important decisions were made which eventually hampered the whole project. Project failures typically involve a number of faulty decisions, but within those many mistakes, some specific key decisions are the generators of the sequence of proceedings that eventually lead to catastrophe.

Key Decision 1 – An alteration in tactics

At the beginning of a project strategic decisions are made that set the project’s track. In the case of DIA, a tactical blunder was made that caused “flip-flop” to be made part way through the project.

Before requesting for bids for an integrated system in the middle of 1991, the DIA’s Project Management team had expected that individual airlines would take care of their own baggage handling activities.

In 1991, the DIA’s Project Management team altered their strategy and realized that an integrated baggage handling system needed to be built. This forced them to take back the responsibility from the individual airlines and carry out the whole project themselves. This alteration of strategy arrived only about two years prior to the airport’s scheduled inauguration date. This timing of the decision played a major part behind the unnecessary schedule stress that the project was exposed to.

Although the decision made sense in the way that only one integrated system would be there with the entire responsibility on the DIA authorities, the timing of the decision was not proper. This led to a delay in start of the integration project. They got two years less time to finish it.

The significant point that the airport’s Project Management team failed to see was that the change in technology required an analogous change in the organizational accountabilities. The failure to identify that change signifies a planning failure that happened during the start of the project.

Overall, the mistake made was a failure to associate the airport’s overall business strategy (the aim of having one of the world’s most competent airports) with the sub-strategy of building the baggage system.

Key Decision 2 – The decision to continue

What is even more surprising is that in spite of knowing that there was insufficient time, both DIA’s Project Management team and BAE wanted to go ahead with the full-scale project.

Before entering into the BAE contract, there were at least three clear suggestions that the project was not feasible in two years’ time:

Breier Neidle Patrone Associates report clearly showed that the complexity was too much for the system to be built efficaciously.

The three bids received pointed out that none of the vendors could have developed the system in time before the opening date.

Munich Airport warned that a much simpler version made for them took about two years to be built and around another 6 months to get rid of the bugs.

BAE initially did not choose to bid for the project, but the Chief Engineer of DIA directly talked with them and managed to convince them by stating the amount of prestige that was involved with such a large scale project.

Many from inside the BAE raised their concern regarding the complexity of the system to be developed and the lack of time. But all the expert advice was ignored and they went ahead with the project with a development time of 2 years.

Many factors may have led them into that trap and likely issues that may have influenced the decision making. Other than the Chief Engineer’s perspective and BAE’s interests there were other factors or beliefs which made the whole project start.

Both sides would have acknowledged that they were working within a constricted timeframe and the pressure to move rapidly might have caused them to put due-diligence to one side.

The inherent belief that such a large airport would not function effectively without an automated system. As a matter of fact, the airport is functioning effectively with a manual system in place.

Key Decision 3 – The timeline, budget and scope

Deciding on the timeline, budget and scope of the project is a critical issue and even more critical is committing on them to your customer. BAE did exactly that to the DIA’s project management team. The decision to give a firm promise to scope, schedule and budget transmitted substantial risk onto BAE’s shoulders. This clearly shows that the top management of BAE was not at all aware of the amount of risk that they were handling.

BAE and the DIA’s Project Management team made an additional mistake during the consultations. They excluded the airlines (who were key stakeholders) from the negotiations.

Excluding stakeholders from discussions of key project decisions is always a trailing strategy.

Key Decision 4 – Acknowledgement of the alteration requests

BAE and DIA’s Project management team could not escape from the stakeholders’ pressures. Although they decided during negotiations that no change requests would be entertained, they had to accept them as the pressure was on them to meet the stakeholder needs. The stakeholders in this case being the airlines, which they ignored during the original negotiations.

Some of these requests forced them to make significant changes in portions where they thought work was already completed.

Incorporating these changes had other troubling repercussions. They failed to realize the effect these changes might have and how they would increase the complexity of the whole system.

Although some people thought about the effects, their voices did not seem to have reached the higher decision making authorities. There was big communication disarray.

Key Decision 5 – Treading an alternate path

A public demonstration of the project was given to the press sometime in 1994 and it was a major embarrassment. It exposed all the flaws of the project and the Mayor immediately ordered for an external consultant to be hired. Mattias Franz of Logplan Consulting of Germany was asked to look into the matter [3] . Based on his report, the Mayor scraped the project and ordered for the building of a manual trolley system at an additional expenditure of about $50M USD [4] .

Although the Mayor took a very intelligent decision, it revealed another major flaw with the project. By the time the Mayor took action, the project was already 6 months behind schedule and had missed a number of opening dates.

The missed opening dates and the tragic demo indicate that those at the top echelon actually had almost no clue about the true status of the project.

A project of such size and complexity should always have an external consultant or expert looking after it throughout the whole developmental and implementation phase.

Some other failure facts

While the inefficient estimation of complexity, absence of planning, fruitless communications and lowly management oversight drove the catastrophe, the project underwent many other difficulties that multiplied the problems.

Some of those issues were inevitable, but others were most probably a consequence of the time crunch the project was facing. Among the additional issues that impacted the project;

Failure of Risk Management

Throughout its developmental and implementation phases the project faced a number of technical problems for which they had not accounted for. These things aggravated their already haphazard situation.

Such problems were likely foreseeable had the team a little bit more attentive on risk management activities. Again possibly as a result of the time crunch under which they were working, suitable risk management tactics seem not to have been developed.

Change in Leadership

In 1992 The Chief Engineer died. He was the system’s de facto guarantor and his death left the project deprived of much required leadership. According to reports, his interim replacement lacked the in-depth engineering knowledge essential to understand the system. The replacement manager also had to take care of his previous duties and it stretched him to the limits.

Issues with Architecture and Design

A number of reports specify that the there was an inherent problem with the design that was chosen. It was unnecessarily complicated and prone to bugs. Some of the issues were:

There were more than 100 individual PCs in the system. They were all networked together. If any one of the PC failed, there could have been an outage, as there was no automatic backup taken of the data.

As the nature of the design recommended a distributed structure, (with PCs scattered around the different areas), it added to the trouble of solving problems when they came up,

The worst thing about the system was its inability to detect jams. So, whenever a jam occurred, it kept piling on more and more baggage and thereby worsening it.

Again time crunch could have been a reason for the design problems. In such a situation people settle for the first design or solution they can think of. That is exactly what would likely have happened. In addition time crunch often forces teams to concentrate on the “happy path” design without spending time on devising strategies to counter the problem or make the system fault tolerant.

Conclusion

The DIA catastrophe is a prototype for failure a lot of other IT implementation projects have followed. As with so many other failures, DIA suffered from;

The inefficient estimation of complexity

An absence of proper planning resulting in consequent alterations in strategy

Extreme schedule pressure

Absence of due diligence

Committing to public and customer in the face of enormous risks and uncertainty

Inefficient management of stakeholders

Communication gaps and collapses

Design not failsafe

Inefficient risk management

Failure to understand the repercussions of change requests

Absence of management oversight

While the above facts denote contributors to the letdown, there is one single problem that existed in the center of it all. For a project to be successful people need to make effective decisions and that requires a number of elements. The main two elements are expertise and knowledge. None of the teams involved in developing the DIA’s baggage handling system had prior experience of a developing and implementing system of this magnitude.

That lack of knowledge, along with the fact that advice from experts was habitually ignored, is the epicenter of the fiasco.

The original planning decisions i.e. to go ahead with a single airport wide integrated system (in spite of being too late to do so) and the firm’s votive commitments to scope, timeline and budget all represented decisions that were made by people who did not possess the required knowledge. The miscalculations resulting from those choices were the sparks that kindled the fire.

Often we have to face situations which we have never faced before and do not know how to proceed without risks. The success or failure of such a situation depends on the way we react to it. The step should ideally be recognizing the situation and its nitty-gritties, but the whole DIA project management team and BAE managers failed to do so. Had they acknowledged their absence of knowledge and the ambiguity they were facing, measures could have been taken to reduce the uncertainty. One of them could have been taking suggestions from experts who had some kind of previous experience in that kind of projects.

The cheerful side of the story is that in Feb 1995 DIA did ultimately open and in spite of using a large manual trolley based system, proved to be a great success [5] . The apprehensions of a manual system being too slow for and airport like DIA and would result in increase in the turnaround time of the aircrafts, was never proved.