The technology perspective
Jim Peters
CTO, SITA
See the presentation that accompanied the speech
(595 KB)
Transcript of Speech:
Greetings everyone, I am Jim Peters, the Chief Technology Officer at SITA. One of my roles at SITA is to lead a dedicated team doing research into new and potentially disruptive technologies. We call it SITA Lab, and it was founded last year. In my presentation today, I wanted to give the Lab perspective on some of the projects we have been working on with disruptive technology.
But first, a quick definition of what is Disruptive Technology. The term was first coined by Harvard professor Clayton Christensen in 1994 in a book called the Innovators Dilemma. Christensen categorized technology advancement into three categories, the first is sustaining technology, which is the normal incremental updates products get, which is the traditional "New and Improved" category of technology. The second is Revolutionary Technology where one technology is developed that replaces the existing technology. A good example would be the replacement of vinyl LPs with Compact Discs. And the third category is Disruptive technology, defined on Wikipedia as "A technological innovation, product, or service that uses a disruptive strategy, rather than a revolutionary or sustaining strategy, to overturn the existing dominant technologies or status quo"
In The Innovators Dilemma, Christensen describes how disruptive technology does not start off as a "mainstream" hit, and in fact, it doesn't really address the current market, and doesn't fit in. It is only after creating new markets and products that the technology then comes to challenge the existing products and markets. Continuing with the music example, MP3 technology for downloading music was initially ignored by the mainstream music industry, but after becoming popular with tools for music swapping such as Napster, it is now turning the music industry on its head.
So to summarize Christensen's Innovator's Dilemma, disruptive technology is both a threat and an opportunity. Initially, it is difficult to invest in because it really doesn't fit the existing market, but if you don't pay attention to it, it can come back and threaten your existing markets and products.
Gartner defines disruptive technology as causing major change in the accepted way of doing things, including business models, processes, revenue streams, industry dynamics and consumer behaviors. Their current top-ten list of disruptive technologies includes:
- Multi-core processors
- Virtualization
- Social Networking
- Cloud Computing
- Web Mashups
- Advanced User Interfaces
- Ubiquitous Computing
- Contextual Computing
- Augmented Reality ... and
- Semantics
The fact that you may not of even heard some of these terms before, does support the notion of disruptive technology as something that develops out of the mainstream, and only gets noticed when it starts to impact the existing markets and products. I will give a couple of examples of work SITA has on-going in a couple of these areas later on, but looking at this list, the question arises, how do we make sure we don't miss the boat, and have these technologies later catch us unaware and unprepared to deal with their potential impact?
I list here some famous examples of historically "bad guesses". I am sure you have seen some of these:
"What use could the company make of an electric toy", this from Western Union when it turned down rights to the telephone. At the time they were the dominant communication company using the telegraph.
There is no reason anyone would want a computer in their home, from the president of Digital, and the head of IBM predicting the world-wide market for computers at a total of 5.
And my personal favorite, Bill Gates predicting 640 Kilobytes of memory as 'enough for anybody'. My laptop has 2Gig of memory - which is 2 million Kilobytes.
The next two are from my own personal experience. In the mid-90s I was building websites for travel companies, and after presenting to a GDS their first website, which at the time was more or less a basic electronic brochure, I pointed out that the underlying technology of the Internet, the world-wide-web and the web browser we very powerful and they could be used for a lot more. The then CEO reacted by saying "This website is the only thing we will ever have to do with the Internet"
And finally, my brother-in-law worked in the mid-90s at Egghead Software, which at the time was one of the dominant retailers of software with over 200 stores in malls across America. He invited me to do a demo of the Internet to the Egghead management team, and their reaction was more or less that the Internet posed no threat to them. They went bankrupt 5 years later.
On the other side of the coin are stories of run-away hype and investment in new technologies that did not work out so well. Pointcast networks came out in 1996 with a screensaver that displayed customized news stories using a new thing called "push technologies", and it was HOT! In 1997 News Corp tried to buy them for $450M and the deal fell thru, but within in only two more years the hype had subsided and the company ended up selling for only $7m in 1999 and the service was then shutdown a year later.
Iridium networks is another example. Between 1994-1998 they launched 66 satellites at a cost of $6B in order to provide satellite phone service. There has been lots of analysis into the reasons why this venture failed, but the upshot was that it was sold in 2001 for $25 million.
One thing to note about both these failures is that the underlying technology has survived and thrived. Push technology and satellite communications are both now considered fairly mainstream. The issue is Christensen's Innovators Dilemma, how to time the efforts to commercialize disruptive technology so as to not miss the boat, which brings us to the question.
What to do? Ignore it? Or Embrace it?
In the end, the answer is not easy, if it were there would not be so many famous failures nor would Harvard professors make money writing books on how to answer this question. But let me give you my view from the SITA Lab perspective.
The SITA approach via the Lab and its innovation groups within the business units, is to first invest in a dedicated group to investigate new technologies and develop prototypes in order to understand the technology potential. As the technology develops, it can be moved from a prototype into Incubation. This process includes doing some trials with customers to see what works and what doesn't. And you have to be open in this process that some things will not work, so we must be willing to accept some failures.
I want to give you a couple of examples of investigations we have underway. The first is around Virtualization and Cloud Computing and the second is in the area of mobility.
Virtualization technology allows you to abstract a software system from the underlying hardware that it runs on. A basic example is the Apple Mac which can be setup to run both the Apple operating system and Windows at the same time and you can switch between them using software from companies like VM Ware.
In the data center, this same technology can be used to put more than one system on a server to get better utilization of the hardware. Better utilization results in lower cost, and this technology really brings new economies of scale. Within the average data center, moving to virtualized servers can result in a 70% reduction in the amount of hardware required.
One place where this technology has been used in anger is in the big internet players like Google, Yahoo and Amazon. What I call here the "mega-Internet Data Centers". According to a study by Berkley University, the levels of advantage these players have on utilization and cost savings against the average corporate data center are between 5-7 times. This means instead of paying $95 per MB of network bandwidth, they pay $13. Instead of paying $2.20 per GB of storage, they pay 40 cents. And they can allocate one administrator per 1000 servers, instead of the normal 140 servers.
One of the results of achieving these economies of scale, is the concept of Cloud Computing, which is the idea to sell this processing, network and storage as a product on its own, so you can put your systems into these data centers rather than running them yourselves. You can get new systems deployed on-demand at significantly lower costs.
Amazon is one company leading in this area of on-demand, scalable cloud computing. They call their service the Elastic Computing Cloud or EC2. They use the term Elastic because they have so much capacity, that they can allow any one user to virtually grab as much computing power as they can use, on the fly, without any need to order and wait for expansions.
This is a screen shot of the EC2 dashboard for systems we are running in SITA Lab in the cloud as part of a program to test out how it works. To get this setup, all that required was to create a user id and enter a credit card number to open the account. After that, you can either build your own software system or use pre-built systems that Amazon provides that do common functions like a web server or a database server. You are billed based on how much processor and storage and network you actually use, and it can expand from 1 server to over 500 on the fly based on the workloads you throw at it.
What opportunities are there for the ATI with virtualization and cloud computing. Data Center Server Consolidation via virtualization is a clear opportunity for cost savings. In the 2009 Airline IT survey it is listed as the number one area where IT is looming to invest in order to save.
Another opportunity is to centralize your PC management by using Virtual Desktops, which allow the use of thin-clients in remote offices and airports and to manage all the user's data back in the data center, rather than having to worry about PC's in the field.
The opportunity to Expand into the Cloud is still not completely clear. While it shows great promise, the reality is that you cannot just start moving your core systems into the cloud. First you must deal with security, backup and recovery, monitoring and management, and compatibility with custom software packages and the cloud platform. These issues still need to be explored and resolved.
One recommendation is to first look at the Cloud as a resource for your software testing. One neat thing about the elastic cloud is that 1000 servers for one hour cost the same as 1 server for 1000 hours. This means you could use the Cloud for doing some large-scale performance testing. For example, before running an online ticket sale, you may want to test that your eCommerce software will be capable of managing the load and you could run a simulation using Cloud resources.
In SITA we are running projects to investigate both virtual desktop and secure cloud connectivity to allow us to use it for software testing.
The second area of research that is a major focus is in the area of mobility. I show here the potential use of the mobile across the customer journey, from check-in to paperless travel, making payments, staying informed or interacting with in-fight-entertainment. With the introduction of smarter and faster handsets such as the iPhone and Google's Android, mobility has become an area of major hype for the airline industry.
One side note. It is not just in the passenger area that mobility holds promise but also for all the staff in the airlines and airports who can also benefit from this technology. Gregory Ouillon's team in SITA's communications business unit are actively working with partners such as Appear Networks to incubate and trial mobile applications for staff to optimize airline and airport operations, and we are work closely with them as we look at developing common mobile technology platforms that can be used for both passenger and staff applications.
Given that Wimbledon it currently underway, I thought it would be interesting to highlight an application that IBM is demonstrating. They call it Augmented Reality and if you remember this was in Gartner's top-ten list. It runs on Google's Android phone and combines a number of mobile technologies including the phone's GPS, a compass and the camera. The demo uses the phones camera and while you are on-site walking around at Wimbledon you can point the camera in any direction and a map is overlaid on the picture showing you what you are looking at. You can use it to get directions to where you are want to go or get more information on the services available on-site.
Yesterday I was discussing this with Nancy Grimes from airportwayfinder.com and the opportunity to use this kind of application within the airport to help people find their way to their gates or find the duty free shopping or other services in the airport they want to access. I admit this is futuristic, but the point is to experiment to discover, "what is possible" and understand how it could be used in the future.
In SITA Lab we did also developed a demo on the Android platform using the GPS to automatically discover your nearest airport and display live flight information. We also tested the camera for scanning a bag tag and querying WorldTracer for your lost bag status. The purpose of these prototypes was to gain a better understanding about how to develop apps for these new smartphones.
What we learned was that there are a lot of compatibility issues between the different handsets such as android, iphone, Windows mobile, Blackberry and Nokia. In effect, it means you have to write each application 5 times, which is going to be very expensive. The other option is to use the web browser on the phone and develop the services as web applications, but the problem then is you can't do all the fancy stuff with the phones GPS and camera, or access the phone's calendar to retrieve your travel itinerary to better understand what you might want to do based on the context of the user.
While I believe these advanced features are definitely going to come eventually, and standards will arise that allow web applications to access the phone's features, for now, we realized we need to walk before we run, and the next step should be basic but a useful set of functions.
So we took these findings to one of our customers, Malaysia Airlines and created a pilot project to do a live trial of some basic functions using the devices web browser. The main challenges for the project is to get it to work on the different size screens and device types that support a good browser. The application includes shopping, schedules, check-in and flight status and will be launched in Q4.
A key design issue is usability, given the small screens of different sizes, meaning you can't just copy these pages from your corporate website and expect them to work. And you need to focus on keeping it simple and clean.
So let me conclude...some key messages to leave you with. First, ignore disruptive technologies at your own peril. You can see from the famous quotes and failures, you need a strategy for managing the threats and opportunities of disruptive technologies. The SITA approach is to track these new technologies and trial them through continuous prototypes and customer pilots.
Finally, I encourage you to leverage the community aspects of SITA, and feel free to bring your ideas to SITA Lab and our Innovation groups, and see how we can do some of this research together.
Share or bookmark with:
Facebook
Twitter
Digg
Delicious
Technorati
