Inevitably, lines and bottlenecks occur. Keeping the operation running smoothly requires careful planning. This is the job of BAA, formerly the British Airports Authorities. BAA is the UK?s largest airport operator and owns Heathrow and Gatwick, two of the world´s busiest airports.

The BAA Operational Research group has been using WITNESS for several years to help airport managers tackle these bottlenecks. WITNESS allows them to build computer-based models of airport operations and interactively modify parameters to analyse the effects of proposed changes before they´re put into place.

Pat O´Grady manager of the OR group, found WITNESS a valuable tool, because it runs on a PC, is visual and allows user interaction. "We were able to go to managers at Heathrow and Gatwick and show them the benefits of simulation. The visual side really caught their imagination. It was very attractive for them to actually see what was happening. They quickly came to appreciate the value of WITNESS. We´re building general models for check-in, security, baggage claim, and immigration. When a manager wants to look at one of these areas, we take the model, insert the specific data and give a rapid response."

Over the past few years, WITNESS has been used to tackle various bottlenecks and problems throughout the airport; ranging from security search to immigration to aircraft parking. Ian Badger, the manager of Heathrow´s Terminal 3, is a frequent client of the OR group. "The airport environment is ever-changing. Facilities that were running smoothly last month, might be inadequate next month. Therefore I need to know what action to take to avoid passengers having to suffer severe congestion and delay."

Security Search at Heathrow

One example of this occurred at security search in Terminal 3 last year. Although all passengers are automatically screened and all cabin bags X-rayed at security search, a certain percentage have to be searched by hand. Following the Lockerbie investigation, the Government announced that the number of passengers and bags that had to be searched by hand would increase by 65%. This led to long lines and poor service at security.

However, before Ian Badger recruited more staff to meet the BAA passenger service standard for security, he asked the OR group to assess what increase in staff was required over the day and whether any savings could be made through changing the layout of the area.

At the end of the simulation project, it was clear that the existing layout was still the optimum one, but each X-ray machine needed to accommodate one extra security person in order to maximise efficiency. The project has also enabled the terminal managers to gain a much better idea of how the passenger flow through the security area is related to the layout and the staffing arrangement. The results were then used as a basis for a manpower planning study

The project has also led to more simulation work using WITNESS. Because the model is very general, it has been used to investigate the optimum layouts and staffing arrangements in the other terminals at Heathrow and Gatwick.

Immigration at Gatwick

Another familiar bottleneck at airports is at immigration, and Gatwick´s North Terminal is no exception. "Many long haul flights arrive very early in the morning, but this is when immigration rosters are at their lowest," says O´Grady. "Since passengers are processed according to nationality and not origin of departure flight, we have a lot of complaints from U.S. passengers who get stuck at the back of lines with consequent long processing times." As a result, the terminal manager at Gatwick asked the OR group to build an immigration model and evaluate alternatives for improving the passenger flow.

Like the security search model, it takes an airline schedule and simulates the wait times and line lengths with different desk manning scenarios. The model can also investigate the effect of additional nationality channels (e.g., putting all U.S. passport holders in a separate channel) or processing passengers according to origin of departure flight.

Although the wait at immigration could only be shortened with the help of the airline scheduling committee and the Home Office, the model was a useful tool in persuading them what action should be taken. "Because of WITNESS, we could show the airline scheduling committee how the lines were likely to build up," says O´Grady. The study showed that there should be a limit on the number of non-EC (European Community) flights each hour and a minimum separation time for those flights from countries with nationalities with long process times.

General Airport Models

"We are now trying to build a set of general WITNESS models; one for security, immigration, baggage claim, check-in and aircraft parking," says O´te;Grady. "So when a terminal manager asks us to look at the bottlenecks in a particular area, we can simply pick up the model, insert the relevant assumptions and provide a rapid response."

At the moment, the OR group is further developing its aircraft parking model. When an aircraft lands, it is allocated a stand ("parking place") based on its terminal of arrival, airline code, size and time of departure. However, if an aircraft is allocated a stand away from the terminal, the passengers have to be transported by coach.

The model contains all the various rules about which aircraft can occupy which stand and allows for variations between the estimated and actual times of arrival. When allocating stands, it minimises the shuttling required and maximises the passenger flow through the stands that have the shortest walking distance.

"One-Off" Models

The OR group has also used WITNESS to look at "one-off" problems. As O´Grady explains "We tend to do two types of a model: the first is generalised, flexible and can be applied to several sites. The second is where we just look at the vertical circulation between the station and departures level at the new terminal at Stansted."

Just before opening day in June last year, Stansted management were concerned that passengers arriving at the station would have to wait longer than the BAA service standards for a lift to the departures level. The BAA service standard is that passengers should not wait longer than 30 seconds on average. However, the OR group was able to respond within a few days by building a small but accurate model. The model showed that, on average, passengers would have to wait only half the BAA service standard and this reassured Stansted management that they had enough lifts to cope with the passenger flows on opening day.

WITNESS: A Key Strategic Tool

WITNESS is becoming more widespread within BAA as a day-to-day planning tool as terminal managers discover the benefits of simulation in managing their facilities more efficiently. This has been helped by BAA?s chief executive, Sir John Egan, who has also been convinced of the benefits of modelling airport operations. The OR group is confident that WITNESS has an increasingly important future both as a tactical and as a strategic decision tool in the planning and operational areas of the business.