In a blockbuster move, American Airlines and HP agreed to abandon their joint Passenger Service System project for the airline, and American hopes to select a new vendor with an existing reservations system soon.
The decision, and details about their proposed settlement, emerged from a June 7 filing and accompanying joint motion as part of American parent AMR Corp.’s Chapter 11 case in US Bankruptcy Court, Southern District of New York. The court is to hear the matter June 28.
In sum, American exercised its option to terminate parts of the agreement, which would have had HP Enterprise Services develop the hardware and software to operate a new reservations system for the airline by October 2014, after the two companies  failed over the last two years “to reach agreement on a number of key issues affecting the project’s direction …” the airline stated.
The parties then worked out a termination agreement, which they are asking the court to accept.
The companies’ Realtime Passenger Services System Agreement, signed February 1, 2010, would have had HP Enterprise Services, which is the successor to Electronic Data Systems, provide the airline with “availability, reservations, ticketing and airport processing,” the airline stated.
It isn’t widely known, but HP actually operates American Airlines’ current reservations system, using Sabre software, and it will continue to do so for the immediate future under the terms of the settlement, if the court gives the OK.
Other terms of the settlement include:
- American Airlines will be allowed “to pursue other options to replace their existing Passenger Service system;”
- HP will continue to operate American’s existing reservation system under the terms of an IT services agreement dating to January 1, 2008;
- American Airlines will not have to pay HP a termination fee;
- HP will have an unsecured claim of $7.65 million for work it has done on the PSS [as of February 2012, HP already was a $30.8 million creditor]; and
- “Significant costly and protracted litigation will be completely avoided,” the airline states.
There is also an important amendment to the termination agreement. In it, American Airlines gets reduced costs and expenses related to HP’s operation of the airline’s exising reservations system, which uses Sabre software.
Also, the amendment provides that American and HP will have joint ownership of intellectual property created during work on the new PSS.
Although American Airlines and HP halted development work on the new reservations system early this year, months after AMR filed for bankruptcy protection on November 29, 2011, the proposed settlement puts a bold exclamation point on their joint project’s demise and sets off a lot of maneuvering for the contract to host the reservation system for a major, albeit weakened, airline.
If all this sounds familiar, United Airlines several years ago enlisted Amadeus to take over PSS functions from Travelport’s Apollo system, but the airline, facing its own financial crisis, backed out of the agreement.
In an ironic twist, given the American-HP news, the merged United-Continental ultimately transitioned over to HP for its PSS.
So what now for American Airlines and its reservations system?
If the court approves the settlement, HP will continue to operate American’s PSS using Sabre software for the time being.
The existing American-HP PSS agreement remains in effect until the bankruptcy court takes action, although is “in a state of suspension,” American states.
“Most importantly the Jetstream [American's name for the new PSS] initiative to replace American’s PSS remains a high priority and is a critical part of our future as a stronger, more competitive and customer-centric airline,” American stated.
However, an American spokesperson says the airline is “moving away from building its own” system and “hopes to make a decision soon” on a vendor which already has an existing solution in the marketplace.
For its part, HP states:
As you know, HP does not comment about our clients’ business or legal situations. American Airlines has been and remains a valued client of HP. We can tell you that HP remains committed to the development of scalable travel and transportation industry solutions that enable airlines to improve business results.
The HP-American falling out has to be seen as a victory for Sabre in the short-term because now its software will continue to be used for American’s PSS for a longer period than the airline envisioned.
Sabre could be a candidate to extend its agreement with the airline, but considering the enmity between Sabre and American, a long-term PSS relationship into the future would be a long-shot.
And, it’s unclear how ITA Software makes out now that American Airlines is shopping for a new PSS. ITA Software had contracted with American Airlines to have the airline use ITA’s inventory control solution as part of the Jetstream project.
The jockeying to become American Airline’s new PSS favorite is undoubtedly under way and will be influenced by merger fever and the outcome of the airline’s restructuring.
Related posts:
Given the amount of effort that went into this design, one cannot help but feel that designing a complete PSS is no longer the right path. We have now a long line of failed PSS projects. Not wishing to embarrass those involved I wont name them, but there are a lot.
Perhaps now is the time to re-think the design of a PSS and look for a less singular approach. Distribution management and customer service tools have improved enormously in recent years alleviating the need for a big bang approach to PSSs. There is a lot of functionality that sits in sub-systems that provide high value to the products and processes of an airline and its stakeholders. (Despite the efforts of many gatekeepers to keep access closed).
However – we are still seeing a need to re-architect the core of the reservation process. Time to discard the Telex one way message based designs of the 1950s.
As an aside comment. Perhaps the industry is ruing the day when IBM abandoned its TPF hegemony. On the other hand this opens up a tremendous opportunity for innovation with the Airline Technology community should be welcoming.
So folks – thinking caps on – what would you do with the proverbial Magic Wand now?
Answers on a papyrus sheet please to….
Cheers
Not a giant surprise, given the lack of progress over the last 12-18 months. As someone who worked on a similar project based on the Unisys AirCore system, I can feel their pain…
I actually think that HP’s “AirSOA” initiative is still a viable proposition moving forward, integrating core PSS components with standalone applications into vertical cloud offerings.
But, as Timothy pointed out, that doesn’t change the fact that nobody has truly succeeded in building a large scale inventory and reservation process core from scratch.
I would love to know what the actual cost of this project was. Certainly not the 7,65 million USD mentioned, given the length of the project and the number of people involved.
What about Amadeus? They are truly the only ones with a new generation PSS that is live.
Carla: American Airlines is in oneworld and Amadeus is the designated common IT platform for the Star Alliance. Could there be a deal? Dicey. Also, Madrid-based Amadeus has never hosted a U.S.-based airline. Amadeus has a deal to handle international for Southwest and will be busy with that implementation for some time.
Amadeus was one of the final two contenders, along with EDS/HP, for the Jetstream project after a five year and mult-million dollar engagement. The issue then is the same as it is now, build or buy , and in their arrogance, they decided they could not accept being part of a community of airlines in which they did not run the show. As Maya said, they believe that they are so unique that an off the shelf system, even one as agile as Amadeus’ Altea wouldn’t work for them. In retrospect, did they make the wrong decision? Yes they did!
How is Amadeus Altea a new generation PSS? Their core reservation platform is still TPF
The fact that TPF is still used for a small number of applications within Amadeus is not the same as TPF being the ‘core reservation platform’. Further, the Inventory, DCS and Load control applications which were of most interest to American for its Jetstream project are built entirely using new architectures.
New architectures? The Application is still client server. Looking at the screens there is a lot of TPF parsing here. I looked at the Amadeus RES system, still to good old green screen. They do know how to sell their products though.
Dennis, you’re correct that Star chose Amadeus for the common IT platform, but Amadeus is also used by many of the cornerstone (no pun intended) airlines in oneworld and Skyteam.
From working at AA, I know they consider their way of doing business unique, but so does just about every airline I’ve met with over the years, including Southwest. Arguably, it’s a far more defensible statement from WN, and if Altea can work for WN, and it already works for so many other of their oneworld partners, then there’s no reason it couldn’t also accommodate AA.
In addition, Amadeus will have completely decommissioned TPF by 2013
Completely decommissioned TPF by 2013? They told in 2009 already that they would stop usinf TPF. I would like to refer to the year 2000 press release when they took over British Airways system, which is still their coore and will remain thier core for the comming years.
Eric: You make great points.
Carla: There won’t be ANY TPFs humming away for Altea after 2013? For real? Hard to believe:)
Dennis: For real!
Sorry, heard this “for real” for many many years. Amadeus’ marketing is known to be so good they hardly need products.
No TPFs at 1A after 2013? Carla, that’s nonsense! There will continue to be millions of Type B messages to/from WorldTracer for interline bags, connections to Timatic, etc. (each month). I can think of 20 areas that will still be connected via 30-40-50 year-old technology well into the 21st Century.
At the time AA’s RFP went-out, no off-the-shelf system was the right approach. That need has changed since Chapter 11.
Maya is brilliant to get AA out of Jetstream during Chapter 11 Re-Org to eliminate litigation costs, etc. Everything has changed in the past 18 months for AA -and- since Google bought ITA. (Who cares that CX, QF, etc., are on 1A — that won’t matter a year from now. Change is coming!) Clearly 99% of AA’s needs are -not- DCS or W&B. Their needs are driven by inventory (to OTAs, their IBE, corporate, etc.), sales and ticketing. They can move to an open system for these and worry about DCS, W&B, etc., later.
Follow the AA/Sabre Direct Connect lawsuit. You notice AA has not wanted that rejected during Chapter 11. That outcome will re-shape the industry (again), not just in the U.S., but everywhere. Cutting $9 from the cost of each PNR/eTKT will be huge!
No, it’s not nonsense. Most of Altea is already running on open systems today. There are a few low-level pieces that are still in TPF, and as Carla already said, they’re set to be replaced by 2013.
Ever see the EDS commercial about building airplanes in flight? That’s essentially what 1A has done with eradicating TPF. It’s been so transparent to the end users and third parties that you wouldn’t know it was happening unless you asked.
Based on the rest of your answer, I suspect you’re confusing TPF with TTY…. The two have nothing to do with each other. Otherwise, the interfaces to Worldtracer, Timatic, and the ability to exchange type A & B traffic wouldn’t be available to Navitaire, Radixx, NetTracer, or the dozen or two ALCS based systems still in operation.
Your view that 99% of AA’s needs aren’t DCS or W&B is also quite interesting…
First of all, W&B was out of scope due to the way AA”s internal architecture works — PSS replacement has never include the components living in FOS. Horizon was always a separate effort, and is still proceeding as far as I can tell.
Second, I was part of the functional team who put together the initial RFP requirements, and evaluated the responses and vendor demos from Sabre, Unisys, and Amadeus back in 2005. Based on that, plus the experience of working for Amadeus on the other side of the process for the past six years, I’d say you’re really underestimating how large a role the DCS functionality plays. Having the best shopping & buying platform around is useless if you can’t successfully deliver on the day of departure.
Lastly… I think you’re really mixing apples and grapefruit by connecting AA’s distribution approach to their PSS decision. In a perfect world, a PSS system should be capable of supporting whatever business model a carrier wants to use. To the best of my knowledge, Altea can do that.
Other systems? Some do, some don’t. You don’t have to look too far to find examples where an airline chose a system based on their at-the-time business model, and wound up hamstrung when their model evolved.
HP’s attempt at building a new reservation system was a complete disaster. They started a start from scratch effort and stood up very high $ local teams. The teams started sprinting with no architecture,no requirements and no project plan. They attempted to use very new technology which cannot be used to build real-time reservation systems eg Oracle ADF.This Agile build was performed for 10 months until Sep 2011.
Management changed and the new thought was if you follow Agile you cannot complete by 2014. HP buys a product Air Core from Unisys. The estimate to extend Aircore for AA was 3 years. Meanwhile AA goes belly up and backs out. HP has spent tons of money and has produced nothing.
HP has two .net products called EMD and TRR. Both of these are very poor quality,non scalable products that cannot be used by larger airlines without major changes.
HP has failed miserably at product development in Transportation industry.
Fools rush in… Dealing with large scale real-time transaction processing is a craft. It requires dedicated, and above all skilled people, a proper design, a well described scope etc. Scalability should be in the genes of these systems (a separation between operating system and application as introduced by IBM in the sixties is a VERY good idea. An operating system that deals with most complexities of large scale processing is not such a bad idea either). Everyone working on such a system should continuously be willing to spend time on shaving off milliseconds in every layer. The danger of doing such a complex project via the ‘agile’ method is that no-one has an overview, resulting in the most expensive spaghetti object bowl in history. Admittedly, a ‘nextgen’ bowl of spaghetti, which could be considered progress. This has nothing to do with ‘new’ or ‘old’ (how old is Unix?) technology. it is about how to use technology. I witnessed a ‘near miss’ project myself, where we were just in time to dictate some old fashioned rulings regarding functional design and operational awareness, get rid of some monstrous object creators and gadget maniacs. Performance could still be saved as an afterthought, but should have been addressed in a far earlier stage. We had a narrow escape… but now run an operation ‘nextgen’ DCS…
Completely agree. HP didn’t have the experience or skill set to perform such massive software product development.
Returning to Timothy’s proverbial magic wand and thinking caps: bad news, no such thing as a magic solution. I say: evolution instead of revolution. Actually that’s what Amadeus did, although probably inadvertently: They added layers of business logic to the existing system 1 software (aka global core). And of course it was enriched with completely new developments in the load control and inventory area’s. This is how you keep the shop open while refurbishing. I think ( and believe have proven) it can be done better (if you accept the fact that a core large transaction processor can actually be an advantage. The TPF operating system is a gem, the problem lies in the outdated applications, and even for these there is a cure ). The answer certainly is not in half baked systems and lots of empty promises, but somehow airlines let themselves be tempted over and over again.