We are told that hapless BA passengers face days of chaos
following computer failure which disrupted operations. It will not just be a simple matter of fixing
the computer and then starting up where they left off. Aircraft will be in the wrong place and their
maintenance disrupted. Crews may have
exceeded their duty time and will need to rest before working again. Passengers will be piled up all over the
place. Chaos indeed which will be a
massive reputation hit to BA (already suffering from the free meals issue). Alex Cruz, the BA Chief Executive said he was
“extremely sorry” and the airline said that the problems were due to “a power
supply issue.” BA dismissed speculation
of a cyber-attack (they would have to unless forced to admit it). Whatever the
cause, there seem to be 2 issues.
Firstly, when the management system crashes, why isn’t there a rehearsed
back-up plan to keep the business going?
Secondly, when massive disruption occurs, why is the reputational impact
compounded by poor customer information and service?
My first RAF Squadron, No 57, based at RAF Marham in Norfolk
(fondly dubbed “El Adem with grass”), was a nuclear deterrent “V Bomber” unit
in the process of re-equipping and re-roling to the air-to-air refuelling
role. The heart of the Squadron was the
Operation Room from where the flying programme (the business) was planned and managed. Management tools in those days were limited
to pencil and paper and data recorded on wall displays. The wall displays were huge perspex sheets on
which was neatly written in grease pencil (“Chinagraph”) all the information
necessary to run the flying business – aircraft and crew location, state of
training, tasks, medical fitness, and administrative duties, together with more
general information about security and alert states. In short, all the information necessary for
the Squadron to go to war. In those
days, we were completely immune to cyber-attack but we did used to joke that
should a malevolent Warsaw Pact agent wish to neutralise the V Bomber effort,
all he would have to do would be to break into the Squadron Ops room equipped with
a rag and can of methylated spirits and wipe the ops displays clean!
As the Cold War intensified during the 1970s and 80s, so did
the resilience of war plans. The concept
of NATO tactical evaluation “TACEVAL” evolved and periodically units would be
evaluated at no-notice on their ability to conduct operations under realistic
wartime conditions. Resilience and
redundancy were repeatedly tested, it being almost a given that the primary operational
management systems would have been “wiped out” at the outset. In 1982, at the Headquarters of Allied Air
Forces Central Europe, I saw the alliance working under exercise conditions at
first hand and pretty impressive it was.
So, my first question to BA would be: why do you,
apparently, not have a rehearsed plan to revert to basics when the automatics
go wrong? Could it be that automatic
systems have become so much a fundamental part of the business that no team of
humans understands the mechanics of the business anymore? Forgive me for
reverting to Marc Goodman’s excellent book “Future Crimes?” He observes that:
“Much of the world’s infrastructure utilise supervisory
control and data acquisition (SCADA) systems to function. SCADA systems automatically monitor and
adjust switching, manufacturing, and other process control activities, based on
digitised feedback data gathered by sensors.”
In other words, command and control systems, SCADA, are not
only complex, they are fiendishly complicated ie difficult, if not impossible, for
humans to comprehend. Computer power is
growing at the same time as our ability to comprehend that data grid and its interconnections
is fading. The cost and complexity of IT
systems propels companies to outsource their problems to 3rd
parties. Whilst this could show
short-term financial saving, in the longer term it merely reinforces management
remoteness from the systems controlling their business. Eventually fatigue will set in and humans
will just give up trying to understand in rather the same way as pilots did
when flight management computers replaced navigators. Again, Goodman points out that Ray Kurzweil
predicts a law of accelerating returns of computer price and performance. Kurzweil says that there will be a moment when
computer progress is so rapid that it outpaces mankind’s ability to comprehend
it. Although Kurzweil says that point
will occur in 2045, he may be underestimating progress. None of this excuses a lack of preparedness for
a failure of automation. However, in the
lessons learned from the current crisis, the cost of training to retain a
capability for manual reversion will have to be weighed against the
compensation costs and medium term (at least) reputational damage to BA.
Turning to the hapless victims who appear to have been badly
let down, it is difficult not to feel sympathetic. Having travelled a fair bit, I have
experienced my fair share of disruption.
However, 2 recent incidents are relevant. A few weeks ago, travelling back to Yorkshire
from Kings Cross, I arrived at the station to find all the relevant departures
labelled “cancelled” or the scheduled departure left blank. There was no other information. There were no tannoy announcements. The display boards maintained their sullen
denials. None of the train companies thought
to send someone into the crowd with a loud-hailer. The whole uncertainty could have been reduced
by a simple announcement on the display board – “get any train North and we
will sort out your ticket later.” In
contrast, last year, we arrived at Schiphol en-route to Bergen. I should have anticipated problems ahead
because, as we left the aircraft, the teenage crew were grinning from ear to
ear having just completed a full Cat IIIB approach and landing – “we didn’t see
anything until the turn off,” said the fresh-faced captain, rather
proudly. We later learned that Schiphol
had been almost paralysed with thick and persistent fog and there were very few
movements in or out. Ominous lines of
passengers appeared and we prepared ourselves for the worst. However, as we were looking for the end of
the queue, KLM employees were exhorting passengers not to queue but to consult
Twitter or Facebook instead. Of course,
not all KLM passengers would be savvy with social media or, indeed have a smart
mobile phone. I classed myself as
relatively house trained so gave Twitter a go.
Amazingly, within minutes my personal plight had been acknowledged by
KLM customer service and, shortly after, a revised itinerary for a departure
the following evening issued. This
enabled us to get away from the airport and find a hotel downtown and, armed
with the certainty of our onward travel plans, able to make coordinating
arrangements with Hurtigruten and our Norway cruise. Later, my claim for expenses was also efficiently
handled on Twitter. Overall, a very
satisfactory customer experience and rather better than the hollow “extremely
sorry” from the Chief Executive.
The secret of KLM’s success is the harnessing of social media.
Having proved the concept with the Icelandic volcano disruption, KLM uses “Salesforce”
technologies at the heart of their operation.
KLM claims to have transformed into a socially connected business that
can intelligently and efficiently collect and track all social conversations on
performance, measure brand sentiment, and most relevantly to the current
crisis, respond within one hour and resolve within 24. That was certainly my
experience. BA may well feel it has
something to learn from Salesforce and, in which case, they should talk to my
talented daughter who just happens to work for that Company!