Blockchain in Travel for Dummies: All You Need to Know – for Now

Interaction with Other Trend Topics Such as Artificial Intelligence

Artificial intelligence (AI) is a term used for machines capable of performing tasks that require intelligence. Technologies trying to achieve this goal include machine learning, artificial neural networks and deep learning. We had our own experiments trying to sort the shopping results for flights into an order like a customer would pick, supported by personalization, and in less than 13 % were we successful picking the flight that was in the end picked. While it can even be argued that to pick the right flight which was later chosen by the user in 13 % is not even that bad – we are not talking about a fifty-fifty chance here – still in 87 % of the cases the user would have been disappointeC.

Meanwhile, blockchain is essentially a new filing system for digital information, which stores data in an encrypted, distributed ledger format. Because data is encrypted and distributed across many different computers, it enables the creation of tamper-proof, highly robust databases which can be read and updated only by those with permission.

I read an article explaining why AI and blockchain are made for each other:

1. Data in blockchains is supposedly highly secure due to cryptography, hence blockchains are supposedly safe to store sensitive data such as PII (Personal Identifiably Information) which is expensive to store safely for large companies today (and still data breaches occur). As blockchains hold their database in an encrypted state, only the private keys have to be kept safe – reducing the amount of data stored in a safe environment immensely. AI is working on building algorithms to be able to work with encrypted data.
2. Blockchain can help humans understand AI decisions. Due to fact that blockchains are recorded one-by-one there may be a way to assess and/or audit decisions AI systems have made.
3. Computers today are fast but not intelligent. Hence, hashing algorithm require large amount of computing power. AI is an attempt to make computers intelligent, hence, they may become dangerous to crack such security algorithms. However, if AI is also included in the blockchain, then it’s a fair game again.

Pros and Cons of Blockchain and What Critics Say

You hear everything from “blockchain is most revolutionary technology to arrive in decades” to the “most over-hyped technology”. Depending on who you talk to, it will change “everything” or “nothing”. Well, one thing certain is, that the truth will be somewhere in between.

Blockchain may be favorable to classical transaction as they are anonymous, have a potentially lower transaction cost and are separated from governments – hence, it is a community and no government can manipulate the ‘currency’ (by e.g. printing more money). However, challenges are the difficulty to exchange such currency into real currency and its potential to be used for illegal activity including tax fraud, which may cause governments to ban them.

Critics also say that after ten years, nobody has come up with a use for blockchain besides currency speculation and illegal transactions and wonder if the reason that nobody has adopted a distributed ledger at scale is because nobody wants it? As an example, payments and banking is provided. Blockchain was to power currencies like Bitcoin, because there was now a costless, instant way to exchange value without the middleman taking a cut. It didn’t take long for that dream to fall apart. For one thing, there’s already a costless, instant way to exchange value without a middleman: cash. When I wanted to sell my wife’s car, the moment I put it up on the internet for sale, a proclaimed ‘soldier’ wanted to buy that car for his father who coincidently happened to be a mechanic, hence be able to deal with the problems the car had. But the only way he was ‘able’ to pay was through PayPal as he was supposedly stationed outside of the country in Germany right now. So, I thought I can trust PayPal and wanted to pursue it, however, luckily, we still called PayPal and found out that it was a scam. PayPal does not secure car sales. So, I guess a person would have picked up the car and title and a few days later, the ‘buyer’ would have claimed they never received our car and asked for the money back (or it would have been transferred from a compromised PayPal account – which apparently there are many – but the thieves apparently cannot do anything with those accounts without raising red flags). We ended up giving the car to the dealership with a significant price drop, but at least we got our money. Our experience with cutting out the middleman would have been short lived.

But even if Bitcoins were considered a substitute for dollars, still Visa and MasterCard actually sit on top of dollar-based banking transactions, providing a set of additional value-added services like enabling banks to track fraud disputes, and verifying the identity of the buyer and seller. It turns out that for the person paying for a product, the key feature of a new payment system –  think of PayPal –  is the confidence that if the goods are not as described you will get your money back. And for the person accepting payment, basically the key feature is that their customer has money and is willing to use it. Now, if you add in points, cash-back, credit lines, free checked bags, insurance and other gimmicks the credit card companies offer, you have something that consumers choose and merchants accept. There are also no FDIC guarantees, reversibility of ACH, and if somebody’s Bitcoin account gets drained because their password got stolen, it is probably the last time this person uses a blockchain.

Other barriers for blockchain include speed (as mentioned above), concerns about an open network and volatility (like the rollercoaster ride of cryptocurrencies).

Finally, a huge threat for blockchain technology is the European privacy law. While organizations are planning different kinds of applications for blockchain such as executing contracts, modernizing land registries, even providing new systems for identity management, there is a big unknown on the horizon: the European privacy law.

The European General Data Protection Regulation (GDPR) says people must be able to demand that their personal data is rectified or deleted. A blockchain on the other hand is a growing, shared record of past activities. A blockchain is distributed across many computers. And a key factor for its reliability is that this chain of blocks (essentially transactions) is in practice unchangeable. Not exactly two ambitions which match very well. And with fines for violation of GDPR of up to €20 million or 4 percent of global revenues, blockchain may all of a sudden become a lot less attractive for large, global organizations.

A blockchain is not made for altering or deleting data. While rewriting data is theoretically possible by creating a new “fork” (branch or version) on a blockchain (if most nodes on the network agree) this is certainly only a possible solution for “private” blockchains (such as the Ripple blockchain which is designed to ease payments between a closed group of financial services providers), and nowhere near a solution for the public or “permissionless” blockchains under no-one’s control (such as the Bitcoin or Ethereum).

The GDPR works well in cloud services models, where I store my Personal Identifiable Information (PII) on Amazon Web Services for example. So, with my contract with Amazon I pass on my privacy obligations to them. With a handful provider this works well, but in a decentralized network having a contract with each node just doesn’t work.

Blockchain in Travel

While it still seems to be in its infancy and models are still being developed as well as hypotheses tested, there is a consensus that blockchain may provide a viable architecture to improve nearly any system that relies on a transaction, like the exchange of information, money or something else – which brings us to the travel industry that heavily relies on transactions.

Blockchain is supposed to provide a high level of security (fraud prevention) since it creates an immutable digital ledger of transactions. Actions can be defined and carried out by smart contracts, which ensure all parties are abiding by the same rules and which automate (and therefore expedite) transactions (such as settlement). And blockchain offers transparency because those involved can see every entry into the ledger of transactions, but at the same time it provides privacy since transaction details are shared only among participating parties.

The latter puts some question marks on my face: is that something all involved parties want? For example, today’s airline yield management is a big secret. Now, if passengers knew the actual number of empty seats on a flight – and I’m sure smart technology on top of blockchain will figure these things out by observing and analyzing public information available on blockchains (“the Kayak of blockchain”), they would probably delay a purchasing decision. On the other hand “only 2 seats left at that price” rather forces a purchasing decision. Or the price of a hotel has been recorded on the blockchain and cannot be changed any longer – is this a good thing or a bad thing? I was told the airlines could create empty blocks and consequently confuse these blockchain ‘scrapers’ – but it is at least something one needs to keep an eye on – at least on public blockchains.

GDSs, FlightStats, etc. – they all charge for data. Consequently, if that data is saved in a blockchain, theoretically these providers would go out of business – this may be even the idea, but the question becomes, who wants to share information for free? Is the idea, that data is actually not free and paid for by every transaction supported by smart contracts? Travel Risk Management firms would have to pay the blockchain to get data about the travelers of a company in order to provide duty of care?

Challenges and hurdles

Decentralized distribution may sound like a great alternative to the currently distribution oligopoly, however, transaction cost will be an issue to consider. As mentioned, blockchains typically require a lot of computing resources given that data is held multiple times and there is significant cryptographic computation to be undertaken. In systems like Bitcoin and Ethereum, this means there is often a prohibitive fee associated with each transaction, which can represent several percentages of the value being exchanged, making them inappropriate for certain use cases.

Scalability is a major concern for travel as well: most existing blockchains (Bitcoin, Ethereum, etc.) have significant scalability challenges. For example, the Bitcoin network can only process about 7 transactions per second. In comparison, Amadeus processes 100,000 end-user transactions per second in peak times. The comparison may be unfair as Bitcoin chose the speed, when they decided that it should take miners 10 minutes to create a new block. Other protocols, such as Multichain can already process 1,000 transactions per second. In order to increase the performance, the intensity of the validation is supposed to be reduced or only portions of the blockchain are validated (“Sharding”).

Distribution would most likely be a public blockchain: This means, unless I understand something wrong, the only change is a shift of the dependency from a provider to a platform. If all the fares are in one blockchain (platform), you are dependent on that platform and what fees will be charged at some point. Is there any learning curve from recent incidents with Facebook? A platform which was supposedly misused to allow hackers to influence the democracy of some nations, and share ‘fake news’. Falsified user generated content could be a threat – even if the platform is ‘owned’ by the community and not by an organization. A positive example that it can work may be Wikipedia – but even in Wikipedia, how can it be assured that all the information is viable and every administrator is immune to bribery? Is the community a better police than the authority?

Travel uses lots of databases, think of traveler profiles, Passenger Name Records and Super-PNR, etc. a (private) blockchain may be an alternative, but I still would like to understand what is being gained compared to a database besides intrinsic auditing acceptability unless one goes for the full blown picture to cut out the middleman and only have market participants who do not trust each other.

Other areas of interest include payments: Many areas of the industry rely on settlements between parties. Consider a hotel booking where an aggregator, OTA and the hotel need to settle cash and commissions based on pre-defined agreements. Today this is an extremely complex process, and the introduction of smart contracts could automate settlements in many areas of the industry including credit- or debit memos of clearing houses such as ARC or IATA BSP.

Finally, integration with existing systems will be another major hurdle. Today it is hard to make blockchain interoperable with existing IT systems. If a hotel booking is made on a blockchain system, how will it integrate with a system that isn’t on blockchain?

Besides the fact that blockchain might not yet be mature enough with regards to performance, energy cost, privacy, and legality, the promise of a new method of storing data that is secure, immutable and decentralized is intriguing and there are plenty of initiatives already underway in the travel industry. Compare the use case of Winding Tree below as a replacement for the oligopoly-controlled travel industry.

Another use case is the use of blockchain technology to manage flight data. Specifically, the creation and storage of information for each flight which is handled by multiple entities, and hence may get out of sync as data is stored on separate databases. The desire is a “single version of truth.” Blockchain is built as a distributed ledger. The idea would be, commercial aspects of ‘who owns the data’ aside, that all have access to all transactions, and those transactions are cryptographically stored so they cannot be altered. Various entities can access and update the relevant data on a blockchain, and those actions are visible to all participants. This means that automated smart contracts control the transactions or data, rather than one entity. A test was done with four airports and one airline and up to 2 million changes recorded. The technology was considered immature for various reasons including ‘user friendliness’ and scalability (up to 500 airlines and 20,000 airports). Personally, I also have doubts of the amount of data created which cannot be altered for such use case and which needs to be stored redundantly on various servers. If a flight is delayed several times, it is maybe interesting when it finally took off and what the scheduled departure was, but who cares about all the numerous in between updates. I just compare it to my email inbox – and such information goes right to trash. But there is no trash in blockchain – is there?

Even the ‘expected to be eliminated’ GDSs have positive things to say about blockchains: Amadeus is confident that indirect distribution is the most cost-efficient solution for all parties to achieve that aim on a global scale. Amadeus is also confident that they bring so much to the table other than a ledger, that they will not be eliminated as the middleman – as well as the agencies won’t. They may however, be one of the users of blockchain technology if it makes sense – but this wouldn’t transform the industry, just giving them a chance to migrate ancient technology into new technology.

Blockchain in Travel: Use Case Winding Tree

The initiative Winding Tree seems to be the first initiative to replace classical (centric) GDS distribution in the travel industry. The Initial Coin Offering (ICO) of Winding Tree which took place in February 2018 brought Winding Tree almost (depending on the conversation rate) $10-$15 Million USD to be used for e.g. development (16,278 ETH by 7,082 backers in total – 13,860 ETH and 6,346 backers during ICO).

The ambition

As many of you probably know, Winding Tree is an initiative to build a decentralized alternative to the current concentrated travel distribution landscape dominated by a “handful of companies”. This fact was long ago already explained in detail in my book “Value Creation for Travel Distribution”. Winding Tree claims that 95 % of the US Online Travel Agency (OTA) market is dominated by Booking.com and Expedia and 99 % of the non-direct inventory for air travel is handled by three GDSs: Amadeus, Sabre and Travelport – which makes it five companies in the travel industry to control the travel market. Besides record breaking profits, the lack of innovation is one of the concerns stating that some still use outdated mainframe computers (announced to be retired though). The dominated market position is stated in various papers and I also touched on this topic frequently:

• Going GUI: The Cost of Multi-Source Agent Desktops
• US Airways Sabre Verdict: No Winner – only Losers!
• Travel Technology for Dummies: What is Full Content?
• The reality behind NDC: Why NDC won’t break down the GDS oligopoly – yet

Another player that generates billions each year in advertising revenue and silently added searching and shopping capabilities through its acquisition of ITA Software to become a dominant position which should not be overlooked.

It is claimed that

• new market entrants are locked out and hence innovation is prevented (although one of the Winding Tree advisors Johnny Thorsen still managed to land a lucrative acquisition with Mezi by American Express), and
• the concept of a few trusted third parties to ensure the validity of transactions is a security concern as history has shown that those Mega companies have failed to keep data secure stating Sabre and Equifax as an example.

The idea

Contrary, Winding Tree will offer a decentralized alternative to GDS and OTA distribution claiming to reduce cost of distribution and allow more flexibility. It is probably more precise to understand Winding Tree rather as a marketplace instead of a platform, where supplier (e.g. airlines) and buyer (e.g. agencies as part of the traveler) come together. The concept is supported by some well-known supplier such as Lufthansa, who obviously love the fact that suppliers will not be charged any distribution fee. The question is what ‘support by Lufthansa’ means? Have they actually written a check or just allowed their brand to be used? Also, I know for fact that Lufthansa always wants to have several horses in the race. When Lufthansa still had a share in Amadeus they loved to have their own subsidiary Lufthansa Systems bid for and carry out the same projects like Amadeus. This was to fuel innovation by competition – eventually abandon one of the two initiatives.

So, while a main attraction factor is that there is no fee charged to the supplier in order to provide their offer to the market place – which means the whole model explained here is turned upside down. There will only be a “minuscule” transaction fee which will be calculated by the blockchain at the time of transaction without any correlation with the fare. The problem I see here is that you are dependent and these fees are not ‘regulated’ and could theoretically end up higher than todays’ transaction fees – and once you are hooked in, it’s always hard to get out again – at a minimum you spend all your integration efforts for nothing. Referral commissions are also possible, which just adds to the insecurity.

The role of blockchain

Transactions Deployment of Winding Tree is planned on public blockchains. Transactions won’t be confirmed by a central 3rd party, but by the generation of new blocks in a blockchain. Each block simulates a container which can contain multiple transactions (e.g. air-, car- and/or hotel booking). This means that each transaction is pending until somebody generates a new block, which transforms that pending transaction into a committed transaction. The question becomes why should anybody have an interest to generate a new block? The answer would be, because he/she will earn money from the transaction fee mentioned above.

As a technical platform it was decided to use Ethereum and the definition of a transaction is based on smart contracts (which are part of the Ethereum platform). A smart contract is basically a script/program, which will be executed once a transaction is triggered. It is probably a smart choice to build on Ethereum, which is besides Bitcoin and Ripple one platform which has a lot of attention, but at the same time offers more potential than Bitcoin.

Using blockchain technology also means, the typical previously mentioned concerns are (performance, etc.) also present. Similar to the Visa example earlier, it remains to be seen if a blockchain can handle the amount of traffic of a GDS. Security is a factor which is still in the air for blockchains. I want to see if the cost of energy for storing each booking (and its history) forever in multiple locations will cause any harm to our planet – maybe this is something that needs to be added to the carbon factor of a trip. At least with a six-digit letter code there is a mathematical end until the same booking code will be reused and thus the old one erased completely (see here) – how is this arranged in a blockchain?

The difference to a GDS

Similar to a GDS, also with Winding Tree, supplier provide their offers in a central platform. However, the transaction is not executed by the provider of the platform (GDS vs. Winding Tree), but rather by the creation of blocks by independent third parties. The confirmation of the transaction and its security against fraud is based on blockchain technology – hence, a decentralized system.

Winding Tree claims to create a platform (as I said, I think it’s rather a market place) which is built from engineers for engineers, which claims to simplify integration. This is great, however, they also say that each travel firm will have to become a software company – which may be true, but Winding Tree will be another intermediary in the mix, unless all the business moves from GDSs to Winding Tree. So, in the end the picture I drew here just became even more complicated. And if Winding Tree is not the only contestant for a future distribution platform (which wouldn’t make sense even in Winding Trees own words as otherwise Winding Tree controls the market and all the reasons above then apply to Winding Tree), there will be several platform one has to integrate with.

International flights have numerous aspects of different local currencies: point of sale or point of commencement (POS or POC) ticket, several airport charges (security, bag, government, embarkation, passenger service, etc.) of origination, potentially several stop-over and destination airport which Winding Tree claims to solve by blockchain technology (read blockchain currency). The only question is how many people today trust today a virtual computer currency?

Summary: Critical Success Factors for Blockchain in Travel

Besides questions asked in this article, and abstract risk factors, such as due to Moore’s law the used hash function may be comprised at some point, there are some critical success factors:

On a commercial side there is the question about the price to store the blockchain (as explained above), as well as, whether or not a decentralized blockchain system is indeed cheaper than a centralized GDS: a supplier may ‘voluntarily’ define ‘referral commissions’ which may be paid once a transaction is confirmed. Bitcoin supposedly has such a ‘feature’ that only those transactions become blocks – hence are confirmed – that promise a commission to the initiator of the block, depending on offer (= initiator of the block = agency) and demand (supplier).

Also, the supported transaction rates of the blockchain technology is still too low, however, this problem could be solved in the near future with added blockchain features (such as Plasma) or change of philosophy (such as Sharding). In the airline industry there might be another challenge though: As one can read in the press, today, that nearly half of all airline traffic comes from bad bots. How will blockchain deal with this – especially when the energy consumption is already at the borderline? Such ‘bad traffic’ and potentially even ‘bad storage’ distributed all over the network may even increases such energy consumption beyond its limit of being carbon sensitive.

Final Words on Blockchain in Travel

I’m not against Winding Tree or blockchain technology in travel at all! Contrary, I love innovation and new technology. And, the initial situation (an undefined number of market participants who do not trust each other) seems to be an ideal scenario for blockchain technology. My research at this time did not yet fully convince me that blockchain travel technology is ripe, and I’m trying to ask the questions which bother me and may bother others – not to turn the idea down, but rather to get answers and understand solutions. I would love if the audience discussed these topics here on this blog. I personally know most of the advisors of Winding Tree and have done business with several of them in their past roles. And we still support several of their recent initiatives today. Simply the fact of this profound knowledge aboard makes me believe that this must be a success story. I just don’t want to only have my emotional feelings guide me, but also understand the risk and success factors of a blockchain technology in travel on an analytical level. I’m looking forward to hearing all your questions, answers, more concerns and solutions.