Connecting the DOTs — Tech buzz words and Travel Industry
To set the record straight let me start off by saying what I feel — tech for the sake of tech is just a tool to waste time and dumb the intellect. On the other hand, technology that has the possibility and capability of either creating new ways of doing business, over hauling archaic business models and creating new kind of jobs for humans is, if not then pretty-close to, the culmination of human ingenuity.
As any era of business, we are in the middle of one that is saturated by buzz-words. Our business is not an exception to it. Travel as an industry has found its newfound love — technology. Almost every company whether in the leisure, online, business, experience based travel is trying to build capability and rebrand itself as a “technology company in the trade of travel”. Some genuinely are, some are making a genuine effort and some, I would go out on a limb to say this, are just riding the buzz-word wave.
As someone who is passionate as well as circumspect of how technology can have an impact on this very traditional industry of travel, I have been thinking of trying the link the dots and lay them out for some time now. So finally here it is.
Let’s begin with the easy exercise of listing some of the “game changing disruptive” technologies that we have been hearing of from travel customers, service providers, technology companies, ‘the intelligentsia’ et all. Here we go (confined to most relevant to travel industry):
1. AI — everyone’s favorite
2. Machine Learning — oft repeated
3. Deep learning – few understand it
4. Blockchain — the Jedi of buzzwords
5. Virtual Reality – is not only about gaming
6. IoT — the holy grail
7. Natural Language Processing & Automated Speech Recognition
There are more, however for our thought experiment, I believe these will suffice quite well. Over the next few hundred words I intend to achieve two things:
A. Give a brief rundown of exactly what do these technologies mean, so that it sets a good context for our next objective.
B. What and how can these technologies possibly impact, disrupt the existing travel ecosystem as we know of it today.
ARTIFICIAL INTELLIGENCE
AI refers to computer systems or programs that are capable of solving problems and training themselves to solve problems that require — human intelligence and capabilities (e.g. learn, infer, communicate & make decisions) OR require elevated human intelligence capacity (e.g., reviewing huge number of pages of information in minutes)
A quick 9-minute video (courtesy Cold Fusion) beautifully embellishes the definition and meaning of AI here — https://youtu.be/kWmX3pd1f10
At a very high level, today AI applications can be grouped across two dimensions:
1- Narrow AI or WEAK AI — comprising of intelligent systems that are built to perform one task specifically and do it better than humans repeatedly and consistently. Most of today’s AI applications fall under this categorization.
2- Generalized AI or STRONG AI — comprising of intelligent systems that possess the same intelligence as humans and are performing any human task. More than 90% of experts believe this form of AI will not arrive for another 25 years or is never going to be.
Beyond this general categorization AI comprises of multiple sub disciplines, like:
a) Machine Learning — Programs that help algorithms to self-learn and improve through data inputs rather than human coding. Think of how your email spam filter gets better over time to detect and place spam messages.
b) Deep Learning — Sub discipline of Machine Learning that uses big data sets to build models that try to emulate inner learning functions of the human brain. Think of a child learning from her environment and interactions. Deep learning tries to create artificial “neural networks” (the wrinkly part of the brain associated with neural pathways and thoughts). These artificial neural networks are capable of learning from data such as images, video or text, without introducing hand written codes or human domain knowledge.
c) Computer Vision — Deep learning powered programs that can identify object, people in images.
d) Natural Language Processing (NLP) — Programs that can understand written text.
e) Automated Speech Recognition (ASR) — Algorithms, that can recognize and translate spoken language to text.
So where does Travel and Tourism figure when it comes to adoption of AI across firms and the overall industry. Well as per an analysis done by McKinsey Global Institute’s Artificial Intelligence, Next Digital Frontier? discussion paper based on a survey of 3000+ AI aware C-Level executives across industries, Travel and Tourism ranks at the bottom of the list.
Artificial intelligence — application and possibilities for Travel
Before we get into the interesting stuff on where AI is being used or can be potentially used in the travel industry, lets do a quick refresher of the industry demographics. Broadly speaking the industry has been categorized into two heads — Business Travel and Leisure Travel. The lines continue to blur, however these two categories of travel, based on the objective of travel and who is the generator of the travel request, still remain relevant today. Beyond this very broad classification the travel industry can be broken down into Online travel platforms (e.g. Egencia, Booking.com etc.), traditional experience based travel shops like Cox and Kings, brick and mortar business travel houses like American Express GBT, Carlson Wagonlit Travel etc., sharing economy platforms like Air BnB, crowd sourcing travel platforms like Trip Advisor, meta search engines like Kayak.com and finally the Airline companies, Hotel chains and GDS companies like Sabre, Amadeus etc.
Recommendation Engines — Probably the most widely used test case of machine learning is recommendation engines and fare prediction algorithms used by online booking agencies like Expedia, meta search engines like Kayak and even some corporate travel houses that are trying to use prediction algorithms to forecast air and hotel prices.
Chat Bots or Virtual Assistants — One of the oft repeated application of AI, by the industry, this is a very clear and present implication where AI can be used to make servicing more streamlined, standardized and intelligently efficient. The chat bot revolution is being led by well-known leisure travel names like Expedia which uses the Facebook messenger as the platform for its chat bot and Booking.com which has developed its own bot which can be used within its web or mobile version. KLM is one of the airlines to have exposed its customers to chat bots. It uses the bot slightly differently where it lets the user’s Facebook messenger plug in to its checkout page. The bot is available in 13 languages and in a typical week responds to more than 15,000 social conversations in a dozen different languages.
Business Travel though is a different story. Chat bots still need to make their mark felt in any real sense of the word across this spectrum of the industry. Which is quite ironical as bots based on AI, NLP and ASR seem to be the ideal solution for an industry that is based on tenets of servicing well, servicing quick and servicing consistently. Cost as an overhead is a constant Achilles Heel for any call center. In a typical business travel call center at least 60% of reservation requests are for simple point-to-point bookings that need to be completed within a set frame work of corporate policy. A seemingly simple problem for an algorithm to solve. The big question mark is the ability of a chat bot to efficiently use NLP and ASR to understand text and voice requests to translate into a set of search commands and fulfillment functions. Not to mention, the pain of providing after hours coverage to business travelers wanting to make changes or explore options at night. A work-life paradigm that chat bots will have no complaint against.
Deep Learning — I personally feel that deep learning will have a huge impact on travel as we know it today. The very essence of deep learning is to mimic the human brain function to make the program self-aware and provide context aware reasoning, prediction and answers. What better space for deep learning to make a mark than travel where the entire value chain is made up of a multitude of interdependencies, each of which can have an impact on the outcome of the travel transaction — traveler experience. In other words, how does a traveler react to certain stimuli or outcomes? Deep learning can help create immensely accurate personalization algorithms for hotels and airline. Navigate and manage duty of care programs. Even create prediction tools for loyalty programs that could calculate probability of customer stickiness or attrition.
BLOCK CHAIN
Ok so in its most simplified, non-technical sense a Blockchain is essentially a shared database with a ledger of transactions, asset ownership or any other activity that with an audit trail. Then how is it different from any database? Well, a few features make blockchain more robust and interesting:
1- Its decentralized. Which means any transaction verification and authorization is carried out without a middleman or a clearinghouse.
2- It instead relies on a network of participants with permissions to validate transactions.
3- Its completely distributed. Meaning each participant or party own digital replicas of the synchronized ledger.
4- It is super secure as it uses cryptographic mathematics to authenticate.
5- And its “immutable”. Transactions are time stamped and tamper proof. With processes to correct but not update or delete records.
Away from crypto currencies, blockchains are being experimented to create test cases across a plethora of industries like, supply chain, job and education records, asset registry and tracking, trading and settlement, identity management (very interesting concept, more on this later!).
Blockchain — application and possibilities for Travel
The big questions — Can and how block chains be used in the travel industry? Are they being used now, as we speak? There is research underway and some companies committing to block chain test cases in travel, nevertheless nothing significantly big as Hyperledger or the Ethereum Enterprise Alliance has yet made a splash in travel. However, two domains within travel are potential kick starters of blockchain disruption — Content Distribution and Identity Management. ShoCard (a digital identity and authentication platform (21 employees) and SITA, the IT company for Air Transport have been looking for ways to store and manage traveler identity on the blockchain to manage identification and authentication. All of us have been through the hassle of crossing borders when travelling. Imagine a Single Travel Token that stores all your travel documents and identity data on the blockchain. They are encrypted, hashed and anyone (immigration, airport authorities, airlines) can authenticate them using the public key you share with them. Possibly through a simple QR code. No check in lines, no immigration horror stories to say the least. Authorities in America won’t have to place their trust on immigration procedure and protocol in China before letting a traveler in. All that data is stored on the blockchain.
Another interesting space to look out for is the content distribution application in travel. Today this is largely done through two sources — the GDS or through a supplier’s own site. GDS’s play an important role as the platform provides aggregation from across hundreds of airlines and millions of hotel properties which then can used by online platforms like Expedia or traditional business travel companies like Amex GBT, CWT, BCD etc.. As one of the key features of a block chain is that it negates the use of a middle man and makes the process decentralized, the content distribution process as we know of today in travel is something that potentially has the maximum amount of threat or opportunity (depends on how you see it!) to build a test case for blockchain for travel distribution. Winding Tree is an interesting blockchain company that is partnering with Lufthansa, Air New Zealand and Netherlands based Citizen hotels. It allows airlines and hotels to publish available inventory to customers without needing systems that aggregate data on flights and rooms, and could therefore allow them to avoid the fees they currently pay for the use of such systems. Early days for now, but how companies like Winding Tree shape up over the next couple of years might force the industry to start looking at a potentially new business model, where access is more important than ownership of content.
Internet of Things — IOT
Wikidepdia defines IOT as — the network of physical devices, vehicles, home appliances, and other items embedded with electronics, software, sensors, actuators, and connectivity which enables these things to connect and exchange data, creating opportunities for more direct integration of the physical world into computer-based systems, resulting in efficiency improvements, economic benefits, and reduced human exertions.
Expansion in connectivity, mobile adoption, cloud computing, machine learning, NLP all have made IOT a possibility. Not to forget the decline in Sensor costs. Underlying that, the most important factor behind the resurgence of IOT has been the increased need for business to have visibility to data, insights and diversification to new business models.
As an example, Michelin solutions has diversified its revenue stream by growing what it calls “tyre as a service”. Michelin gathers data from RFID chips placed in tyres and then telematics to provide services focused on decreasing tyre expenditure, fuel consumption etc. The tyre related services & solutions unit generated 1 Billion Euros for Michelin in 2016, as reported by the company, and plans to double it by 2020.
IoT — application and possibilities for Travel
The possibilities are unlimited when it comes to IoT application to travel. From personalized services in hotels and flights, to measure the health of passengers through sensors embedded in an airline seat.
What would be much more interesting to see is a change in business models using IoT as the backbone of such models. For example, hotels charging for the number of hours stayed in the hotel room or what amenities were used in the room while your stay (including air, yes that’s measurable too!), instead of check-in / check-out dates. Airlines possibly giving chunks of air miles to organizations and travel management companies for usage instead of a per ticket cost. IoT has given us the ability to create distinct packets of service or modules and the ability of businesses in travel to create revenue models based on such modularized service is what will create the next wave of disruption in the industry. Uber and AirBnb showed us how unutilized assets can be used to generate revenue. IoT can track, monitor and generate insights on usage like none before, and the ability of travel suppliers to use this to create new business and revenue models will potentially differentiate the winners from the laggards.
