|'Legacy system' The 'brick' phone above manufactured |
some 20 years after the NATS computer..
THIS STORY IS LIVE: BOTH MONITORED & UPDATED: RETURN OFTEN
+++ Saturday 13.12.2014 12 Noon: STATEMENT Swanwick controller workstations. In normal operations the number of workstations in use versus in standby fluctuates with the demands of the >>>>> traffic being controlled. In this instance a transition between the two states caused a failure in the system which has not been seen before. The failure meant that the controllers were unable to access all of the data regarding individual flight plans which significantly increases their workload.+++
In our view this is an astonishing admission to make. Clearly as old ie legacy computer technology (some dating from the 1960's - A Comet 4C was the jet of choice) is inside the workstations and in the 21st Century still being used, it overheats and the need to 'preserve' the ancient kit means that when not in use it should be 'in standby' mode.
We noted earlier, the visual displays are akin to CEEFAX or Prestel generated images possibly still reliant on TV style cathode ray tubes rather than flat screen digital LED HD displays which are 'cold running' and consume very low voltage. But the situation at NATS could be worse than we've painted, as we now hear via THE REGISTER that disgraced services firm SERCO provide some of the technical support, and the suggestion is that an incorrectly coded flight plan subsequently embedded on both servers might have caused the computer crash.
Its also been revealed that NATS runs on the 1960's IBM’s mainframe architecture S/390, deﬁned in a time when assembler programming was predominant and compilers were in their childhood. S/390 systems include 31-bit addressing mode, instruction-dependent address formats, limited availability of address displacements and immediate literals, and condition code handling.
In the early 1960s IBM deﬁned the System/360 architecture designed to serve a whole family of systems. The difference the distinguished systems of that family had was the way the instruction set was implemented. The System/360 architecture deﬁned 16 32-bit general purpose registers, 4 64bit ﬂoating point register and a 24-bit address space.
Shortly afterwards, virtual addressing was added to the architecture.
In 1970, System/370 was introduced, providing an enhanced instruction set.
Around 1982 370/XA brought 31-bit addressing.
In 1988 370/ESA introduced support for multiple address spaces.
In the 1990s the ESA/390 architecture was introduced; subsequent machines added over time the relative branch instructions as well as the IEEE ﬂoating-point instruction set.
Effectively the need to turn on and off the Swanwick controller workstations is a major safety issue. An example is any modern office using 21st century computing the units are rarely turned off, if they cease to function, the kit now module based is simply swapped out, with no downtime or loss of data.
+++ Saturday 13.12.2014: 150 flight cancellations confirmed.
+++ EC2612 Passenger Compensation Rules do NOT apply
+++ Swanwick technical failure UPDATE 20:00hrs 12.12.2014
Following a technical fault with the flight data system used by air traffic controllers at Swanwick, NATS can confirm that the system has been restored to full operational capability and a thorough investigation is continuing, to identify the root cause.
Although operational restrictions applied during the failure have been lifted, it will take time for flight operations across the UK to fully recover so passengers should contact their airline for the status of their flight.
We apologise for the impact that this issue has had, and the delays and inconvenience caused.
Simon Hocquard, Director Operations Strategy:
We all know that SESAR is the key to unleashing efficiency in the European ATM system – what NATS is now doing is making that happen far more quickly. Our Deploying SESAR programme is a huge part of our investment programme, but we are fast forwarding it to deliver all the benefits of greater capacity and fuel efficiency, much faster than originally planned. It will allow us to accelerate deployment of SESAR concepts as they mature, on a single operational and technical platform without the constraints of legacy systems. It is the most exciting operational transformation in our history – and it is already under way.
Contrast the two statements above - they came directly from the sold off NATS (National Air Traffic Service) website. NATS manages en-route air traffic services for both civil and military aircraft flying through UK
airspace; NATS Services is the commercial business, providing air traffic control services at 15 UK airports and in Gibraltar..... It is a public-private partnership formed in 2001 between government, holding 49% plus a golden share, and the Airline Group, which has a 42% share. The stakeholders include British Airways, easyJet and the Universities Superannuation Scheme. NATS staff hold a further 5% stake and LHR Airports Limited 4%. The title NATS harks back to National Air Traffic Control Services, established in December 1962 and when some of the technology introduced then is still in operation in 2014!
However returning to the statements above show that with commercial pressure now being applied both by politicians and a commercially driven Board hell bent on meeting service levels to release contractual financial bonus payments, the pressures on Air Traffic Controllers - the humans at the heart of the machine, are being let down - clearly unacceptable.
The core legacy system dates back to the late 1960's and the software still retains elements from those early pioneering days, but when air traffic volumes were far less. But like the 60's flight details are still to this day initially written to slips of paper and slotted into a try of wooden blocks.... the video display system used within ATM still harks back to the golden days of CEEFAX!
NATS in the UK tracks some 5000 aircraft a day and then 'safely' hands those civil aircraft over to other controllers as the flying asset moves out of NATS controlled airspace into say France. It has to be said the French also cause many additional and unwarranted problems for NATS by the number of wildcat strikes or other industrial action incidents they generate.
There is no doubt that when NATS was sold off, the purchasers overpaid for the outdated physical asset. The implication at the time was that the entire system needed to move away from paper slips and wooden blocks and invest in a brand new state of the art computing system. Still hasn't happened, have a look at the skillsets of the Board members below, for a possible answer. You'd clearly expect pointy headed computer people to be at the top table...
The reality is that it is still running old legacy computers needing climate control handling and every time a software upgrade or more correctly a software patch is introduced, there is a real possibility of catastrophic system failure, not matter what the 'spin doctor' at NATS would have you believe. They are putting aircraft and passengers at risk. Any deviation however small, from a recognised schedule involves an element of risk.
The failure of the system on the 12th December is not just a technical problem, or a 'glitch with the software' but an international very public disgrace, as is the brevity of the apology via a website, from NATS for putting the airways at risk.
The problem also is that some media commentators (mainly retired pilots, who had little contact with core operations but were in effect customers of the system) are attempting to suggest that safety was never in question. That is totally wrong, on the TV News, the dire state of ATM NATS style was played out.
One passenger gave a very lucid account, her flight heading for Heathrow was diverted to Paris, then stacked, then taken out of the stack and sent to Brussels, again put into a stack, then finally directed to Heathrow its original intended destination. Total time in flight 3 hours, for a scheduled 80 minute flight.
Now some Budget airlines have reduced the amount of reserve fuel they carry - lighter planes = larger profits, luckily this particular aircraft had enough fuel on board to burn off during the stack process. To add to the distress, passenger compensation under EC261 Passenger Compensation Rules isn't available as the incident was not caused by the airlines, however they face a bill of many millions for missed flights, hotels for in-transit clients, and road/air tranfers for passengers diverted to European Airports.
As for baggage, who knows where it is? A game of eventual catch up involving couriers will now take place. Some commentators estimate extras costs totalling £5 Million. Add another couple of million and you could be close. Diverted aircraft sitting at the wrong airport just clock up parking charges and one off refuelling surcharges....hopefully pilots have their gold Amex cards handy!
NATS say they have back up systems at Swanwick, where the computer complex is sited. The back up is however not a new computer but yet more legacy computer kit loaded with endless updates and patches. They do have back up generators and UPS (Uninterrupted Power Supply) so data can be transferred from one system to another but not seamlessly as up to date computing would permit.
The NAT humans still have to put into play the aircraft reduction process: Plan B if you like, and indeed yesterday the whole of UK Airspace was effectively and very suddenly cleared and closed. It is little wonder that the career of many controllers is somewhat shorter and other will 'burn out' such is the relentless pressure of the job, not helped by legacy hardware, and a Board seemingly sitting on its hands.
A serious high level - independent of NATS - investigation needs to be held immediately without the interference of the European Commission and the composition of the Board also needs some attention. The following this Saturday morning should be hanging their corporate heads in shame....
Some of the directors are 'long in the tooth', doyens of a bygone aviation era. A bit like Concorde still capable of flying, but the financial liability grounded them. Apply the same logic to the NATS main Board Directors, who must collectively carry the can for yesterdays catastrophic computer breakdown, or 'cock up' as it was described by Channel 4 News. They should go. That of course is up to the shareholders.. And as to fixing the glitch? They probably should just turn the computer off and back on again. Works for me..
JULIAN BRAY 01733 345581, Journalist, Broadcaster, Aviation Security & Operations, Travel / Cruise Industry Expert, EQUITY, NUJ, Broadcast ISDN 01733 345020 SKYPE: JULIAN.BRAY.UK e&oe > Updates are on the Website