Consider these imponderables. How long is a piece of string? Is there life on other planets? When will my airline finally return to profit? Most airline managers will have something to say on all these topics. But raise the issue of what information technology implications the year 2000 date change holds for the industry and many executives will shrug their shoulders and refer you to the computer department. Maybe they would take a closer interest if they realised that AMR Corp reckons the problem will cost it at least $100 million.
For years companies in all sectors have adopted an ostrich-like stance with corporate heads buried firmly in the sand. 'This is a general management matter not an IT matter. If general management don't do anything about it, it won't get fixed,' says Robin Guenier, executive director of Taskforce 2000, a non-profit organisation set up to increase awareness of the issue among UK companies. The fact that this is the only government-initiated body in the world underlines the lack of action.
The potential disabling effect of the millennium date change on software applications or systems, which use chips with date codes embedded in them, has been well documented since the early 1990s (see box). Indeed, one leading carrier is understood to have identified problems with the millennium date change back in the early 1970s, but the only action it took then was to 'patch' them. The real problem facing managers in many businesses, including airlines, is that the potential magnitude of the problem has taken too long to filter through to senior management.
This situation has arisen largely as a result of the nature of the IT industry. Computers are seen by users as solution providers and any problems in the IT world are assumed by the uninitiated to be fixable through a program that simply has to be written and then run on a machine. But in the case of the date change no panacea exists that can cure all the ills inherent in legacy systems over 30 years old. Each program has to be audited and checked for 'year 2000 compliancy'.
Establishing whether the software or system is compliant is quite straightforward. 'If you approach this issue at the IT level it is incredibly simple,' says Taskforce 2000's assistant executive director Tom Glasson. But the process of managing the task and implementing the changes, which have to be done by hand, is not so simple. 'I think there is a large percentage of companies who are aware of the problem, but have underestimated the extent of the problem,' suggests Tom Cook, president of Fort Worth-based Sabre Decision Technologies (SDT). Cook is heading up the implementation of the 2000 compliancy programme for American Airlines, Sabre and the other AMR companies.
Industry-wide awareness of the issue is certainly on the increase but Susan Thomas, director of the Worldwide Team 2000 Program at Unisys, which advises carriers such as Ansett, Qantas, Air France and SAS, estimates that some 50 per cent of the world's airlines are 'nowhere' in their year 2000 planning. Indeed, Thomas, who attended an Iata meeting in September on this issue, says she knows of at least one major carrier with a reputation for sound management that was shocked by what it learned at the meeting and admitted it hadn't even started tackling the problem.
And here the problem really starts. Unlike any normal business project, the deadline for implementation of a solution to this issue is not moveable. But such is the shock when managers do realise the size of the task ahead that Guenier says he has had to grapple with the suggestion from more than one senior manager that the year 2000 should be delayed. Indeed, all the businesses that are implementing projects are aiming at a completion date of January 1999 to give themselves a full business cycle in which to test that the reconfigured systems are fully functional.
Moreover, the IT business is very bad at keeping to deadlines. Guenier points to a Sheffield University study which found that 80 per cent of all computer systems projects in the UK were behind schedule, while in the US this figure was 87 per cent.
Then there is the daunting issue of cost. As Guenier points out, there is no visible bottom line benefit to be gleaned from implementing a year 2000 project - apart from the fundamental issue of survival, of course. 'This is going to be very expensive and there are few benefits. The only case to be made is that the company will survive after 2000.'
Cost estimates are almost impossible to come by. Thomas puts a rough estimate of $2 billion on an industry-wide solution. AMR is the only company in the sector to have gone public with an estimate, which Cook confirms is currently running at around $100 million. He adds: 'I feel fairly sure it's going to be north of that neighbourhood. But cost is not an issue really because we have no choice.' The size of American and Sabre means that the AMR estimate will be at the top of the scale, but the lack of hard numbers means that when asked to approve budgets, boards are going to be faced with approximate figures, at best.
This itself highlights how far behind companies are. Glasson quotes a recent multi-sector survey by IBM which found that 80 per cent of IT managers are aware of the problem but that only 18 per cent had a budget to tackle the problem. 'No budget, no plan,' he adds.
Indeed, Northwest Airlines, which Thomas of Unisys describes as being 'extremely well positioned' to tackle the problem, has yet to get have its budget approved, although this is expected by December. Jane Erickson, Year 2000 project manager for the airline, expected to finish her report on what systems 'it is worth fixing and which need upgrading' before the end of November.
Erickson has been working to audit the carrier's systems since January this year and says she has found 'a little over 300 systems that need changing.' Just to illustrate the scale of the task, 25,000 desktop computers worldwide needed checking. The corrective part of the project at Northwest, which includes the crucial issue of testing the altered systems, will go into full swing next January over a period of 24 months.
One major issue that could affect any carrier only starting down the road is the lack of qualified programmers worldwide. Erickson is keeping some of her internal capacity free to pursue strategic IT projects, such as electronic ticketing. Northwest has hired an extra 120 programmers, but still intends to outsource 'a lot of our corrections.' US consulting group Gartners estimates that external suppliers will only be able to cover 40 per cent of the work over all sectors globally.
Chuck Barnhart, SVP applications development at Galileo in Denver, highlights another issue: supplier compliancy. One of Galileo's first acts when its millenium project was initiated two years ago was to introduce a compliancy clause into all its contracts. Indeed the issue of supplier compliance must be the largest single variable in any project as suppliers to airlines come from all sectors, from aircraft manufacturers to food, and from fuel suppliers to banks.
The question of the effect of the date change on the safe operation of the world's airline fleet has produced quite a number of scare stories in the press. Indeed, there is some speculation that at least three carriers plan to ground their fleets on 1 January 2000. But Michael de Jong, manager commercial systems at avionics supplier Smiths Industries, dismisses any such suggestion. He has identified only one area that is date sensitive - the navigational database program, which is updated on a 28-day cycle - and is working to ensure compliancy.
One of the main hurdles facing the airline business is that no-one has a true measure of the extent of the problem. Iata is conducting a survey and hopes to have some preliminary results by the end of the year. But Barry Tate, head of Iata's Year 2000 Taskforce, acknowledges that because of the average response rate on surveys this may still not paint a very clear picture. At the very least he hopes that sending out the questionnaires will increase awareness.
But it's not all doom and gloom. As Cook at SDT points out: 'There is an opportunity for airlines to replace their legacy systems with new solutions rather than updating 20-year-old ones.'
Any airline managers who are still only contemplating the task before them can at least draw comfort from the thought that they are in good company. The results of a survey by the US Congress of all federal departments and agencies published at the end of July found that Nasa, 'one of the most innovative, advanced and computer dependent agencies in the Federal government, has not prepared a plan to solve the problem and does not anticipate having a plan completed until March 1997.' Nasa, it appears, is not alone. The Department of Transportation took an even more ostrich-like approach and failed to respond to the survey.
Source: Airline Business