The airport sector has generally been sluggish to adopt remote data and cloud services for its pricey, server-intensive IT operations, but the early adopters are beginning to feel the benefit. Gatwick Airport in the UK has been pushing hard to scale down its data centre commitments in favour of a suite of cloud-based services from a range of providers, including the likes of Amazon Web Services, Microsoft’s Azure platform and the Box remote data storage service.
Contracting cloud service providers is part of a wider IT programme at Gatwick that aims to empower staff through simplification – employees can now access common applications through their own smartphones and tablets – and customers through new information and navigation services like the Gatwick Street View application launched in November 2013.
Gatwick’s chief information officer Michael Ibbitson has been driving these changes since his arrival in May 2012; we caught up with him to discuss the benefits, challenges and economics of cloud-based airport integration, as well as how cloud computing might play into Gatwick’s proposal for a second runway.
Chris Lo: When you joined Gatwick, what were your initial thoughts on improvements to the airport’s IT infrastructure?
Michael Ibbitson: Well, I think we need to go back a bit further in that story. When GIP [Global Infrastructure Partners] took over the airport in December 2009, they had to separate all the old IT systems from BAA. That’s everything – HR, finance, procurement, all of those types of things. They really hadn’t had the chance yet to get involved in the transformation work that was essential.
When I arrived, fortunately the back-office separation stuff had been done, so that allowed me to focus on what needed to be done in the next four or five years to start transforming the passenger experience with technology. We already had two and a half data centres on site, and there was a need to either reinvest in those or look at an alternative strategy. And when you’re trying to make sure that the passenger experience is at the heart of everything you do, you don’t want to be investing millions of pounds in data centres; you want to just use some of the other services that you have available in this country to do that, such as cloud computing, or remotely hosting data centres, or using things like AWS [Amazon Web Services] or Azure. Then [you can] focus that money on improving the passenger experience.
CL: Did your previous experience working at airports in Abu Dhabi and India influence the direction you are pursuing at Gatwick?
MI: Yeah, definitely. What I saw in those countries was, partly because of the quality of the telecoms service provision there, we couldn’t really go down the cloud services route, so we ended up having to manage things in a very much geographically-centric view around the airport. But I always felt that was quite costly and wasteful, because a lot of airports use the same integration.
That suggested to me that there was an economy of scale where if you have good telecoms service provision, if you build that integration into cloud services or over the internet, you don’t actually need much of that data to touch the airport campus itself. So that was the strategy we decided to run – how much of the integration needs that we have can be built off the airport, and therefore be used multiple times by multiple airports, giving us an economy of scale?
The journey from check-in to security at Gatwick Airport is lined with advertising that does not just simply promote the stores.
CL: You’ve partnered with a number of cloud services providers – how does Gatwick’s system work with all these separate services?
MI: It’s about the overall architecture. Each one has a role to play, but nobody can fulfil all of the requirements, right? We’ll talk a bit about a few of those. We use the Microsoft Azure platform as an integration layer to allow us to move data off the campus. It’s integrated directly with the on-campus enterprise service bus, and therefore allows us to move data into other cloud service tools, whatever they may be. That’s really the role that Azure plays for us today.
We worked with Amadeus to build a product called Airtilus, which is our ACDM [airport collaborative decision-making] portal; ACDM is a Europe-wide initiative to improve the efficiency of airports through integration.
That’s enabled us to provide a web-based portal that all the airlines and ground handlers can log into and see the real-time status of all of their flights and the position where they are in the turn, and it also gives a look-ahead on which flights are potentially going to be delayed and how they can be turned around faster to try and make sure they leave Gatwick on time, even if they’ve arrived delayed. The ultimate result is that the passenger experience is improved through this integration, because what could have been a delayed flight, we now get an alert two, three hours in advance so the ground handler or the airline knows what time limit they’ve got to turn it around. It’s making all this integration come together to improve the whole network.
CL: In what ways has the cloud streamlined the workload for Gatwick’s staff?
MI: The big difference is that the infrastructure layer just becomes invisible to us. Historically airports have had to hire a lot of staff or outsource infrastructure management to other providers, because of the number of servers and the size of the data centres.
Gatwick had something like a thousand servers when ownership transferred to GIP, but now we have significantly less than that – in the hundreds – and we continue to drive that number down. And those servers take a lot of management; you’ve got operating systems, you’ve got patching, you’ve got security issues.
But if you move into a cloud services environment, or a software-as-a-service environment, that infrastructure layer becomes entirely invisible to our operations. For example with Airtilus, we don’t actually have visibility of how many servers that runs on and how they’re kept up to date and managed, because it’s not relevant to us. What’s relevant to us is that the software continues to run inside the agreements of availability that we have.
CL: You’ve previously talked about the scalability of cloud services – how does cloud computing fit into Gatwick’s proposal for a second runway?
Airport-technology.com lists the ten busiest airports in the UK based on passenger traffic in the first eight months of 2013.
MI: When it comes to the second runway, obviously you have to imagine that in theory it potentially doubles the size of Gatwick in terms of passenger numbers, so all of your IT systems have to scale appropriately. When you use a service like Azure or AWS, it’s much easier to scale the infrastructure that you’re running on – things like databases sizes – because they’re all provided as a service that’s agnostic to us. And if your applications are architected correctly, you can ensure that the infrastructure is designed to scale as the application grows in capacity requirements.
And certainly in something like our central search area, where we do a huge amount of biometrics and queue measurement work to make sure the passengers are always getting through security in less than five minutes, time and transaction speed on our enterprise service bus and our databases is absolutely essential. In a world where we have an additional terminal and an additional runway, you can scale those services up quite easily if they’re hosted in the right infrastructure.
CL: Was it tricky to reorient the IT team and the wider staff to this new way of working?
MI: I think the answer to that is that it’s a journey, isn’t it? First of all, our IT team had to start thinking differently about how they wrote RFPs, how they thought about technically architecting the solutions. They had to get used to the idea of not knowing where the hardware was and how it was running.
For the rest of the staff across the business, we moved from a company that was running 400 Blackberrys to bring-your-own-device, and within six weeks we had 1,600 other devices on the bring-your-own-device network. And we can see that continuing to grow. We’re encouraging the staff to actually go and find software-as-a-service solutions for their problems and then bring them to us, and we’ll see how we can integrate them into the rest of the IT architecture. We see that starting to happen; it’s not perfect yet, but it’s a journey that will take a number of years as people learn to self-serve a bit more with regards to their IT services.
CL: Does this process have a knock-on effect for the customer service that staff can provide to passengers?
MI: Oh, absolutely. Now that we have bring-your-own-device and we encourage people to bring smartphones and tablets, that means anybody you walk up to in the airport who has a smartphone should know how to get to our website, show you how to subscribe to things like an e-mail flight update and help [passengers] during disruption by getting information to their smartphones via the Yammer application to tell them what’s going on.
With our queue measurement system, we alert the staff in the security area via e-mail to their smartphones now when they should open or close lanes to manage the passenger flow. So from the on-the-spot customer service to the controlled and managed operational benefits, we try to encourage all of that.
CL: Are you also hoping to give customers more options to access services using their own devices, like Gatwick Street View?
MI: We were looking at the idea of whether we wanted to have a Gatwick application that had maps and flight information on it, or whether we should utilise some of the other tools that are out there, like Google Street View and Google Maps. We came to the assessment that our application requires a lot of our development time and a lot of effort, and it has to be downloaded by people so it requires a proactive stance from the customer. Something like Google Maps and Street View – they’re already available on everybody’s smartphone, at least Android and iOS anyway, which was about 85% of the market at the time.
One of the things we’re quite passionate about here is making sure people get through the airport in the least stressful way possible, so they get through to the departure and then relax. And when you have visibility of how your entire journey’s going to unfold from your desktop computer at home before you’ve even set off, you feel a lot more confident when you arrive that you know where you’re going, you know where your gate is. The ultimate knock-on effect is hopefully you plan to spend a bit more time at our restaurants or spend some money in our shops.
CL: What are the economics of implementing a system like this, when compared to the cost of traditional IT infrastructure?
MI: Not having to invest large amounts of capital money in data centres and servers obviously has a benefit, but you do end up with an increase in your operational costs. Historically in UK airports, we had what we call RAB-based regulation, which was about capital expenditure.
Now that Gatwick has moved into a new contracts and commitments framework with its airlines, where we have commercial contracts, the capital expenditure model has changed, and actually, for us, the operational expenditure model that we’re moving into, where things will be more based on passenger numbers and aircraft movements, makes sense. It means that our IT costs can move up and down to more accurately reflect the current traffic or economic conditions, which will obviously work for us. So the model is a lower CAPEX spend – we still have some, obviously, we have networks to build, something to connect these devices to – and a slightly higher OPEX. But overall it comes out at a lower cash spend.
CL: What’s your roadmap for continuing to push forward this programme over the next few years?
MI: We just launched Airtilus for its first use on 17 June, and it’s now in full use in our airfield operations building and by our flow leads, who are the people who try to ensure the aircraft arrive and depart on time. We’re rolling that out to the airlines in the coming months and we expect a couple of hundred users on that system throughout the airport community by the end of this year. So that’s quite exciting.
What’s also exciting is integration with the airlines that we’re working on, to make sure we have a much more real-time understanding of EasyJet’s network and how they’re operating, and the same with British Airways. We’re working closely with Norwegian as well now that they’ve started their US flights, to make sure that we can try and boost their efficiency and productivity, and make sure that those routes are a huge success.
And then we’ve got this huge North Terminal development; we’re going to refurbish the North Terminal over the next two or three years. One of the key areas there, is implementing self-service bag drops to try and eliminate queues from check-in. That’s a really interesting and important step forward. And of course, repeating the innovations we’ve made in security, biometrics and queue measurement in South Terminal for the North Terminal over the next two or three years as well. There’s a really fascinating programme of work going on.