Support Skift’s Independent JournalismMake a Contribution Now
Editor’s Note: Last month Skift started a new three-part series on the digital and mobile transformation of transportation, the changing face of cities, and how they both have the potential to upend everything we know about travel.
The first article in this series introduced the mobility era and how new mobility providers and aggregators created new transportation products and services in cities all over the world.
The second part, along with the big 170+ slides presentation by the author, is below and well worth a deep read. You will find all articles in the coming weeks here.
This “revolution” only started a couple of years ago, with the creation of public transportation standards and the “open data” movement. Then advances in technology and computing power were leveraged more broadly to offer services such as real-time multi-mode (car, taxi, bus, train etc) mobility journey planner.
The Need for Open and Live Public Transit Data
With the advent of mobile computing, users of public transit are now demanding instant planning based on live mobility public transportation providers information. Where’s my bus? Should I take a taxi, a bike or use public transportation now?
New solutions were obviously needed beyond the basic and proprietary public transportation app and website, especially to offer a unified search solution. In order to offer such a seamless multi-mode of transportation journey planner, transportation data was needed.
Open Data Is the Fuel of Smart Mobility
Open transportation data started to be released by mobility providers, city by city, around the world.When data was not available the citizen crowd provided it to small mobility aggregators (Waze is may be the best example so far).
Data could be then shared, mapped, crunched by algorithms to propose immediate value-added predictive or prescriptive services. Value started to be generated by new entrants, with open data, in three ways: by making data available, by making data accessible and by making data valuable “in context”.
Making data available means that you should first think about what data to provide, but also to fund the platform to be used and finally, what license to apply to the data.
Share-alike license, for example, requires copies or adaptations of the data to be released under the same or similar licence as the original. Of course, providing public transport data through portal and APIs is a cost, but the most difficult part today is still to force cities to “open” their data.
Sometimes, governments had to pass a law to force open the fair exchange of this types of data (like in France). Another way of doing it is to modify all the contracts done by local authority with mobility providers, to provide open data by default (like in Paris). Nevertheless, getting taxi or black cab live data is still an open topic in all major cities. The giants of the sharing economy adore people sharing their services, but do not share their data with others easily.
Making data accessible is generally done via an open data portal or via APIs. Recently governments like in Germany and cities like London started trying to trade these data sets.
With the digital world being more and more data hungry (especially with the Internet of Things development and for the future driverless car), trading data will be a major business in the future, for anyone that could ensure a certain level of data quality and “freshness”. We could also imagine, for a group of independent hotels located in the same city, to publicize their fares and availability through open data instead of going through a property management system or an online booking service.
Making data valuable is now a top priority on all smart mobility stakeholders’ agenda. Sometimes data alone is hard to exploit and need to be enriched, depending on the targeted context. For example, you could merge the schedule of flights with the statistics of their average delay over years, couple it with the weather forecasted, and provide it as a unified API.
Valuable data sets are used by algorithm for better decision, but are also leading some clients to go premium. In a world of free, full of advertising, more and more people will be ready to pay for better data, no tracking and a better delivered service. For developers, all over the world, these commercial opportunities are the best incentive to build a successful, mobile and digital only, data-driven business. Empathetic services and applications are being invented and offered to the public, in particular to cover their last mile needs, and to reduce their city life frustrations and loss of time.
Think about it, how much are you ready to pay to gain 30 minutes per day?
Two Standards and the Crowd
Google has entered the fray, since it wanted to be able to offer transportation information to its users. It created an open transit open, named General Transit Feed Specification (or GTFS) as an open protocol and provided an API to use it.
GTFS “feeds” allow public transit agencies to publish their transit data and developers to write applications that consume that data in an interoperable way. GTFS-realtime, an extension to GTFS, is a feed specification that allows public transportation agencies to provide real time updates about their fleet to application developers. This has meant the start of the creation of a new app ecosystem, leveraging that information.
In Europe, an another standard was created. The Service Interface for Real Time Information (SIRI) is an XML protocol allowing distributed computers to exchange real time information about public transport services and vehicles. It was developed by the European Committee for Standardization, with the initial participation of France, Germany, Scandinavia, and the UK.
At the same time other outsiders, not willing to wait or trying to force transportation providers to publish their data, started to gather data using crowdsourcing. Scout, Ototo, Moovit and others were able to quickly map certain cities and have started offering alerts using user-generated content.
Lastly, public authorities have started to “force” by contract or by law (like in France) to publish transportation data. The Pandora Box has been open, and the ecosystem is eager to re-invent mobility applications and services.
The Importance of Maps
The last barrier to innovation was maps. Most of the time results of a trip journey planner were presented on a map. Google Transit was one of the first to offer a dedicated transit map with a journey planner. The high cost of using such maps was considered by the ecosystem as an issue, since it grows with popularity and the number of requests sent.
Open Street Maps (OSM), created by Steve Coast in the UK in 2004, followed the Wikipedia model and took some time to cover most of the world. OSM is supported by the OpenStreetMap Foundation, a non-profit organization registered in England. For some years now, OSM is considered as a very good alternative to Google Maps, or Microsoft Here, even if some black holes subsists. Mapquest offers both open data and “licensed data” (updated more regularly) and Telenav Scout app is using OSM for GPS navigation.
One of the key part of the maps API is the geocoding capability, that is, being able to map a physical address to a map as precisely as possible. If the element to localize is a moving target, like a taxi or licensed car sending its position every five seconds, you have to be able to quickly find the nearest to a given position. Hailo for example, released as open source their own geoindex implementation. The approach they took was “to split the surface of the earth into a grid, each cell of which is a square of fixed size in km”. Uber bought DeCarta, a company specialised in geolocation and mapping. Each second counts.
The commercial race is now on building high fidelity meter-resolution maps able to be used for geocoding, and hopefully these could also be used for the autonomous car, forecasted to be commercially available in 2020. Is this a reason why Google made Google Earth Pro free to use for all recently? We’ll see soon.
The Race for the Best and Quickest Journey Planner Algorithm
This revolution around transit journey planners remained unnoticed from the travel industry. GPS systems had already existing for years, and Waze, created at the end of 2001, was already starting to map the world with their crowdsourcing solution. The true revolution came — besides the open data movement and the GTFS standard — from the race to build the best public transportation algorithm, one that could compute transport solutions quickly.
It was a well known problem, and Dijkstra in 1959 published the first algorithm concerning the creation of a “Tree of minimum total length between n nodes“. But we had to wait until 2005 and a challenge organised by the Center for Discrete Mathematics and Theoretical Mathematical Challenge at Rutger University to get algorithm that could be delivering results in a usable way. In 2008, Robert Geisberger proposed the ” Contraction Hierarchies ” algorithm, used today by Open Source Routing Machine.
Travel booking startup Rome2Rio developed its own algorithm after having studied several academic papers, including Transfer Patterns from Google published in 2010, and selected a particularly fast driving directions algorithm based on OpenStreetMap data.
And in 2012, Daniel Delling from Microsoft Research proposed Raptor, the algorithm which is now among of the most used today. The algorithm “engine” and code were released as open source, fueling again the mobility ecosystem with solutions for one of the most complex part of the journey planner.
From Transit Apps to Multimodal Door-to-Door Journey Planner
Once public transit data, maps and journey planner algorithm were available, the ecosystem started innovating. For example, Navitia.io, a subsidiary of SNCF (French train provider) and Keolis (French public transport provider), released their full stack code as open source. In a matter of weeks, several applications were built on top of it.
The other most well-known open source solution is OpenTripPlanner (OTP), created by a non-profit technology organization called OpenPlans. OTP manages walking, mass transit, bike sharing, and car sharing on an open source platform that cities and individuals can help design.
Rome2Rio, made by by two ex-Microsoft engineers, popularized the door-to-door approach and the multi-modal planning, at least among the early adopters. Planning to go from point A to B was as simple as submitting two real physical address (and not only name of cities or airports like any other online booking travel site), and getting back as results various types of transportation solutions between those points, whether it is air, train, taxi, or any other type of transit. The drawback is that the tool only offers planning, not booking. On the corporate side, KDS Neo and RouteRank, moved further by providing fares and booking capabilities.
The Age of Platforms
Platforms always win at behavior change. Platforms can bring together the combination of people, data and technology that can deliver the right information at the right time in order to influence behavior. As Ming Zeng, Chief Strategy Officer at Chinese platform giant Alibaba, said “It takes three years for a platform to take shape, five years to get initial traction and eight years to hit critical mass at scale”.
The mobility era is the era of digital intermediation offered through digital platforms. Establishing a platform in the center of a robust digital ecosystem is key to succeed. AirBnb, BlaBlaCar, Moovel are all based on rock-solid digital platforms.
At times, the innovation is the platform by itself, letting entrepreneurs build new businesses on top, like A-mano, CitizenGate, or CitizenData. Europe as usual has a project for that, named Simpli-City, a next-generation European-wide service platform allowing the creation of mobility-related services as well as the creation of corresponding apps.
In order to grow their business and win the cities mobility market, mobility aggregators have to build a platform. They also need to offer a seamless user experience through mobile applications, and a unique pervasive payment system, sometimes backed-up by a plastic mobility physical card.
Daimler built the Moovel platform and bought RideScout app; BMW and Keolis invested in Moovit; Apple bought HopStop, etc. The latest kid on the block, PlannerStack, is offering as open source the whole stack for building a journey planner as a platform by integrating all stakeholders from the value chain. You can also use the platform as a service, and they will provide hosting and data integration for you as well.
Platforms enable comparison and transparency: for example real-time rideshare comparison will be possible directly and not only through ad-hoc services like UpHail. Platforms will enable new innovations to be integrated more quickly, especially in the on-the-go payment system like HopOn that proposed mobile payment through ultrasonic sound waves, or NFC for Apple Pay and Google Wallet.
In the End, Open Wins
Open data, standards and open source technologies were used to fueled innovation and to create a fast and growing ecosystem in transport. On the other hand, the traditional travel industry as we know it today is still based on proprietary data, closed APIs, and no real de-facto standards exist today.
In the light of what we’ve seen in transportation, one could then understand the stakes around the airline industry-backed IATA NDC initiative. Sooner or later, the list of hotels by country and city, the list of airline schedules, the taxi position in the city will be provided through open data and API.
The City SDK API is already offering a CitySDK Mobility API and a CitySDK Tourism API. In 2013 Lisbon was the first city to implement the tourism API, followed quickly by Amsterdam, Helsinki, Lamia (Greece), and Rome.
Who noticed these early innovations in the travel industry? Others will be developed in the future, but the big question is who will be the best equipped to leverage them? Part 3, the last in this series, will talk about it.
About the Author: William El Kaim was till recently the Technology Marketing Director of Global Product Innovation at Carlson Wagonlit Travel in Paris, where he contributed to the invention of several travel products and filed a patent. Currently on sabbatical, he specializes in technology innovation in the transport and travel industry. He holds a Ph.D. from University Pierre and Marie Curie (Paris, France) in software engineering.