By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
1987-90: In the beginning
The Civil Aviation Authority (CAA) presents its case to government for a new air traffic control centre and system. The operational date is set in stone as 1996, by which time it is said that the London Air Traffic Control Centre will have reached the end of its life.
The management strategy for the project is defined during 1989 and 1990. However, the Nats project team estimates a minimum eight-year programme - two years longer than agreed.
1990-1: A warning is ignored
An independent consultancy, Mitre Corporation, engaged to assist with the scoping of the project, estimates that it will take 13 years to build a fully-operational system of the type envisaged, giving an operational ("O") date of 2003. Yet the project is authorised to proceed on the basis that it will go live in 1996.
An audit report finds later that National Air Traffic Services (Nats) senior management was keen to demonstrate a "can-do" mentality and was unwilling to retreat from the 1996 date.
By the end of 1991 a competitive system definition contract is let to IBM and Thomson CSF. The New En Route Centre (NERC) is needed to "yield a 40% increase in capacity".
1992: Definition phase falters
IBM and Thomson have largely exhausted the £1m given for the project definition phase. They are unwilling to invest further significant sums themselves. With a substantial number of requirements still undefined, a systems implementation contract is let to IBM. IBM enters into the contract without firm offers from subcontractors which include Siemens Plessey, Logica and Frequentis.
Based on IBM's RS/6000, AIX-based architecture, on a Token Ring network, the system comprises about 200 workstations, displaying both aircraft data and radar information. However the requirements have been set by a small group of air traffic controllers who are not considered as representative of them all.
1994: Computer Weekly's warning
Computer Weekly warns that there are serious system problems but the Civil Aviation Authority denies this.
US company Loral acquires IBM's air traffic division and conducts a secret audit of the Swanwick systems, revealing 21,000 defects in nearly two million lines of code. The whole NERC project, including the Swanwick building, is said to cost £350m. For years MPs assume this to be the total cost of Swanwick, when in fact it is only what is drawn from capital funds.
1995: Acceptance tests halted
Although the system works well on 30 workstations, it proves impossible to scale it up on the full operational configuration. By November acceptance tests of the New En Route Centre (NERC) system are halted.
A quarterly progress report from Nats to the Department of Transport states that "NERC will be online to meet the traffic demands of Summer 1997", although it is already well known to the Nats project team and to the contractor that there are potential problems in the schedule.
1996: Problems become apparent
Details of the problems emerge in public. The contractor submits a plan showing a system acceptance in August 1997. Nats rejects this and says acceptance must be by June 1997 with December 1997 as the revised "O" date.
Lockheed Martin acquires Loral. By the end of 1996, the December 1997 "O" date is put back to March 1998.
Despite the original fixed-price £132m IT contract, the sums paid to Lockheed Martin rise ultimately to £337m. Of the 160 individual payments to the contractor only 15 are for software originally specified. The rest is for additional options taken up by Nats after award of contract. The IT contract is nearly half the total cost of Swanwick.
1997-8: MPs hear evidence
In September 1997, Nats approves a new "O" date of November 1999.
In March 1998 Computer Weekly staff are called to give evidence to the Commons transport committee. We argue for an independent audit of the Swanwick systems, against fierce opposition from Nats. The committee calls for an audit which gains government approval. By October the Government announces that the "O" date has been put back to the winter of 2001-2.
1999: Audit reveals huge cost rise
In June 1999 a confidential paper for a Nats board meeting refers to an audit by Arthur D Little of the finances and management of Swanwick.
It warns, in essence, that the audit has found that the cost to completion of the project will be £180m more than has been announced so far.
As well as revealing that tens of millions have not been included in the Swanwick costs, now about £700m, the auditor's report attacks the culture which "inhibited open debate of potential problems in the project within and outside Nats".
2001-2: Swanwick goes live
In 2001 Nats installs well-written software releases which reduce drastically the number of bugs in the Swanwick system from hundreds to zero. Fourteen years after the CAA first presented its case for a new centre and system, the operational date of January 2002 is confirmed.
27 January 2002: The Swanwick safety case is approved by the Civil Aviation Authority and the NERC becomes operational.
As many European centres are facing major problems with air traffic control systems, the UK is one of the first to have a "new" system in operation.