DB & Google Transit & The Best Laid Plans Of Mice & Men

Rome2rio CTO Bernie Tschirren and I have just returned from a couple of weeks in Europe; we visited Amadeus at their grand campus on the Cote d’Azur, attended the ITB conference in Berlin, and met with a number of companies in Paris. On the in-between weekend Bernie headed to The Netherlands while I visited friends in Italy.

All that travel was made easier by journey planning on Rome2rio — no surprise there — but we also took the opportunity to see what our friends at Google would suggest for our travels. And that, as they say in the classics, is quite a story!

There was quite a stir back in 2012 with the announcement of an exclusive data sharing deal between German Railways (DB) and Google. DB defended that decision in an open letter, saying (forgive the translation if it’s not quite perfect) “The quality of information for our customers is our top priority here.” By that, we assume DB meant that the schedule data must be displayed carefully and in line with various protocols that ensure customers are always seeing accurate, timely information. Restricting distribution of the data to a single partner, Google, was DB’s way of making the data more readily available to consumers while at the same time maintaining control over the quality.

Or so they thought. Check out this Google Transit result for “Amsterdam to Paris”:

Screen Shot 2015-03-16 at 4.36.06 pm

Amsterdam to Paris via Cologne and Frankfurt? Eight hours? I don’t think so! I suspect that seeing one of their services proposed for this route is actually quite an embarrassment for the folks at DB. They entrusted their schedule data to Google to ensure just this sort of ham-fisted result wouldn’t occur, and here is their chosen partner doing exactly what they were trying to avoid. Of course the correct result for this journey is the excellent Thalys high-speed train, a direct service that takes just over three hours.

The wildly inappropriate use of DB routes doesn’t end there. For my trip from Turin to Paris, Google proposed a scenic, 22-hour route that included back-tracking to Milan, meandering through Switzerland, a handful of train changes in Germany, and… well, you get the idea. Here’s Google Transit’s suggestion:

Screen Shot 2015-03-16 at 4.38.38 pm

Goodness me! Twenty hours, seven train changes… DB had sought to protect travelers from seeing poor quality information on 3rd party sites, but surely didn’t imagine that its schedules, even in cases like this where they are displayed quite accurately, could form part of such a bad result for users. They could be forgiven for feeling let down by Google, who don’t appear to be holding up their end of the bargain.

Anyway, back to my travels. I took the more obvious solution, the direct TGV service from Turin to Paris, which runs five times each day and really is a delightful, low-stress way to travel from Italy to the French capital. While on board, I poked around in Google Transit a little more. I found all sorts of routes where, rather than display no results — which might be a better option for their users — Google is displaying similarly inappropriate results along with this disclaimer:

“These results may be incomplete – not all transit agencies in this area have provided their info.

While some operators have joined DB in sharing data with Google — OBB (Austria) and SBB (Switzerland), for example — others, including France’s SNCF, have not. Laying the blame on the holdouts seems a little unfair, and is unlikely to convince any that they should play ball. All of this is clearly an argument for more data openness; after all, the sky hasn’t yet fallen in the UK, the Netherlands or Sweden, all places where government has legislated for public access to all transport data.

We hope the industry sees another lesson here: closed, exclusive arrangements deprive consumers of the benefits that flow from open markets. Making schedule data available to all comers, including highly focused and innovative startups like Rome2rio, Wanderio, GoEuro, FromAtoB and others will always lead to better outcomes for consumers than exclusive arrangements with corporate giants. The proof is out there, plain to see.

Rod Cuthbert

Political and geographic complexity in multi-modal search

We have blogged previously about two interesting challenges that present themselves when you get into the nitty-gritty of building a multi-modal search system:

1) Detecting landmasses. Accurate routing needs to be aware of landmasses and islands, and what is connected to what by land, road, ferry and air.This is critical for finding all possible routes between points, and what combination of ferries and flights will make that connection possible.

2) Political borders. Certain borders cannot be easily crossed; suggesting a driving route from South Korea to North Korea is both unrealistic and an embarrassing user experience. Similar complexities exist in places such as Israel, Pakistan, and Afghanistan.

Political data is also important for displaying accurate place names in our geocoder system. For example, North Elizabeth Station is located in New Jersey state in the USA, hence the fully qualified name of North Elizabeth Station, NJ, USA.

This week we launched significant improvements to the accuracy of our internal system for detecting both landmasses and political regions. Our original implementation utilized data from the Natural Earth dataset, however this data was limited by insufficient resolution and some landmasses were missing. We have now transitioned to the more comprehensive Open Street Maps (OSM) planet data.

The difference is illustrated in the maps below of the Puget Sound area near Seattle, with each landmass represented by a unique color. The original data (left) has much smoother, lower resolution coastlines that are missing much of the detail in the landmass shapes. Some of the smaller islands are completely missing. The new OSM data (right) is more detailed and includes the smaller islands.

landmasses

Steilacoom to Anderson Island is an example query that has been improved. The original, low resolution data caused the routing system to suggest a ferry to nearby Ketron Island instead of Anderson Island (left). The new data fixes this problem (right).

anderson-island

On the political front, Cairo to Amman was a problematic query where Rome2rio suggested driving through Israel (left). The more popular Taba-Aqaba ferry route is now displayed (right) as well as various bus and ferry combinations.

cairo-amman

Whilst developing the new technology, Miles on our team also tackled the engineering challenges involved with implementing a system for very fast lookups of this data. Each lookup provides the landmass and political information for a latitude / longitude co-ordinate, and a search on the Rome2rio site requires thousands of such lookups.

Miles learnt a few interesting facts in the process:

  • The OSM data contains 497,040 separate landmasses (that is, land with a closed coastline).
  • 89% of those landmasses have a coastline perimeter of less than 2 kilometers.
  • The longest bridge between two landmasses is the Donghai Bridge near Shanghai.
  • The region with the greatest density of separate landmasses is around Horsey Island in the UK.

Horsey Island

We will continue geeking out on geospatial data as we keep refining Rome2rio’s search results.

Rome2rio Partners With Loco2 To Provide Integrated Rail Fares & Booking

Thanks to a recently launched integration with UK-based Loco2Rome2rio users can now search and select European train tickets on the site. Try this new feature by clicking on the Schedules & Book button that is displayed for many train routes such as Paris to Avignon:

loco2-1

Rome2rio displays several train itineraries with fares and available classes of travel:

loco2-2

Choose to Book At Loco2 and we will send you directly to Loco2 to complete your ticket purchase:

loco2-3

We chose to partner with Loco2 because they are clearly leading the pack when it comes to providing simple, straight forward European rail bookings on a single site. Loco2 coverage is excellent in the UK, France, Spain, Germany, Italy, Austria, Switzerland, The Netherlands, Belgium, and Denmark.

Until recently we did not imagine being able to link to a single site providing rail bookings across all of these networks. We’re thrilled to see Loco2 make great progress in turning this into a reality.

It is not perfect yet, however. Tickets by Loco2 for Spanish, German, TGV and many international routes are issued as Print-At-Home e-tickets and can be easily purchased by Rome2rio users from around the globe. Other rail tickets are sent by post to UK addresses only, or must be collected from a station prior to travel.

We’ll continue to work with partners such as Loco2 to improve this customer experience with the goal that anyone, anywhere will soon be able to use Rome2rio to easily purchase a rail, bus or ferry journey — anywhere in the world.

Rome2rio Announces Concur Collaboration

Rome2rio today announced a partnership with Concur, the world’s leading provider of spend management solutions and services, to bring Rome2rio’s journey planning solutions to Concur users worldwide. With this partnership, Concur customers will be able to plan a complete door-to-door trip with Rome2rio and have the itinerary automatically sent back to Concur as a travel request prior to booking a trip.

“This collaboration reflects a growing awareness among business travelers that looking carefully at all the transport options for a journey can have a dramatic effect on the overall cost and efficiency of a trip,” said Rod Cuthbert, Rome2rio CEO. “Most travelers focus on air and hotel costs, however in many cases air is simply not the best option. In other cases, the ‘final leg’ of the trip may have an outsize impact on trip cost. We enable business travelers to look at different travel options, compare costs, and then seek approval for the overall cost rather than only certain elements,” he said.

Concur release 1

​Screenshot example shows ​taxi (US$220) and train (US$20) options from Narita Airport
to a downtown Tokyo hotel.

 

Concur customers will benefit from Rome2rio’s comprehensive coverage of ground transportation options in Europe. “Travelers in Europe have many transport mode options, but it’s been difficult to quantify the relative costs on a side by side basis”, said Jigish Avalani SVP & GM, Developer Services, Platform Services. “Being able to quickly compare the relative costs and trip durations associated with air, rail, inter-city coach, self-drive and new options such as rideshare provides an unique benefit for our customers and aligns with our vision of the perfect trip.”​

concur release 2

​Screenshot example shows ​various transport options for a typical European business trip.

 

​We expect to complete the integration of the Concur interface for delivery in the first quarter of 2015.

About Concur

Concur is the leading provider of spend management solutions and services in the world, with 20,000 business clients and 25 million users. Through Concur’s open platform, the entire travel and expense ecosystem of customers, suppliers, and developers can access and extend Concur’s T&E cloud. Learn more at concur.com or the Concur blog.

Uber & Bike Share Options Debut On Rome2rio

Sharing economies are on the rise, nowhere moreso than in the transportation sector. With that in mind, Rome2rio users won’t be surprised to learn about the two increasingly popular transport options we’ve added support for this week.

uber

Uber is the major player in the new wave of smartphone-driven taxi and urban rideshare services. The San Francisco-based company is growing rapidly, with coverage in over 200 cities worldwide, all now available in Rome2rio results.

In many of these cities you can choose from the original UberBLACK premium “black car” service or the new UberX “rideshare” service, which is often less expensive than a local taxicab. For anyone who has had an experience with a dodgy taxi driver while travelling in a foreign city, Uber presents a welcome, reliable alternative.

bikeshare

Bikeshare networks have been deployed in many major tourist destinations, with rapid growth over the last few years. We have now incorporated the bikeshare networks of over 100 cities into Rome2rio, including the popular Vélib’ system in Paris with over 1,230 bike stations, and the London network with 720 stations.

Bicycle sharing networks are ideal for short urban trips and for helping with the “last mile” problem of connecting residents to the public transport network.

You will discover both new transport options in a search such as London Paddington Station to Big Ben. In cities where they are present, Uber will also be displayed as an alternative for the final airport-to-destination leg of a journey.

uber-bikeshare-screenshot

Comparing Coverage: Google Maps, GoEuro & Rome2rio Head To Head

Comparison header

Since our launch four years ago we’ve been passionate about making Rome2rio a truly global travel search site. That means having coverage so complete that Rome2rio can help users discover how to get anywhere by any mode of transport rather than how to get most places in countries X, Y and Z by modes of transport A, B and C but only with operators D, E and F.

The way we see it, incomplete coverage is a little like a street directory with a bunch of pages missing. That’s why we’ve gone to great lengths to import as many routes and operators into Rome2rio as possible. We’ve developed web crawlers, data feed importers and manual data entry tools, all managed by our global team of transport researchers. The result is a database covering some 800,000 routes and over 4,700 transport operators.

So, how does our coverage compare to others in this space? How often do we provide usable, relevant multi-modal search results when others don’t? We decided it was time to find out, by benchmarking our service against results from two other prominent sites offering multi-modal results: Google Maps and GoEuro.

The Competition

Although Maps is not yet marketed as a travel product (and generally lacks pricing and booking information) Google have been adding more and more transit feeds, including important inter-city operators such as German Railways and the entire UK network. A tremendously popular product for Google, it’s becoming clear they see these enhancements to Maps as a bridge between the vast search and travel markets.

GoEuro — who recently announced a $27 million investment round — have quickly become a leading player in the European multi-modal space. We like the GoEuro site a lot, especially the crisp, elegant design and slick user experience. This Berlin-based company is growing fast and seems certain to become a favourite with many users in the European market.

The Test Results

The initial test was based on a randomly selected sample of 25 queries from our own search logs. The head to head test results between the three sites are shown below: a green bar indicates the site provided at least one train, bus, ferry or rideshare result; a red bar indicates no result  was shown. A yellow bar indicates a partial score, where only part of the route was covered, or operator information was missing. We excluded flights and driving options as this data is readily available and well covered by all three sites.

coverage-analysis1

Clearly the difference in coverage offered is significant. Rome2rio nailed all the queries in this initial test set, while Google Maps had complete results for 36.6% of the queries, and gave partial information for another 9.8%. GoEuro provided complete coverage for 14.6% of trips in the query set. When the set is restricted to European locations, a fairer measure for GoEuro, their score increases to 16.2%.

Second Round Tests

Recognising that there may be some bias inherent in the initial benchmark—people visiting Rome2rio may have prior knowledge of the site’s coverage and issue queries that are more likely to be correctly answered—we created a second set of queries, this one drawn entirely from TripAdvisor forums. We mined user questions such as “How to get from Athens to Gastouni?“— there are millions of such questions posed on TA, so it was a rich vein. The resulting queries  are quite challenging, as users are more likely to turn to the TA forums when they cannot easily find transport information elsewhere. In this second analysis, we focused only on trains and buses, as neither Google Maps or GoEuro profess to offer rideshare, and GoEuro does not offer ferries.

coverage-analysis2

Rome2rio’s score slipped slightly on this query set, and there was one query (Cortona to Montevarchi) where Google Maps identified an option that Rome2rio missed. GoEuro achieved only 4.8% coverage (and 6.3% for the European subset). Google Maps also struggled with 28.6% coverage. All up, the delta between the three sites was pretty similar for both query sets.

Points Of Difference

In contrast to Google Maps and Rome2rio, GoEuro are tackling multi-modal search using the traditional live API connection model used by OTAs and meta-search sites, and also employed by multi-modal players RouteRANK, FromAtoB and Waymate. This approach provides full pricing and booking information on the initial search. It does, however, limit coverage to operators who offer an API: that’s OK for most rail operators and the larger coach and ferry operators, but not for the vast majority of transport operators worldwide.

This is a key limiting factor, and suggests that GoEuro will be popular as a booking tool for well-known, high-traffic routes within Europe, but less attractive as a general transport search tool or journey planner. That focus seems consistent with their stated goals, and this benchmark result should be seen in that light.

One of the key ingredients for Rome2rio’s comprehensive coverage is our global research team and their manual data collection efforts. In December last year we explained that although manually collected data represented just 2% of routes in the Rome2rio repository, they appear frequently in search results and represent 49% of user clicks. This is largely because our manual content curation is targeted towards important regions and operators. As our repository has grown this breakdown has shifted slightly:

coverage-analysis3

Going forward, we will continue to seek 100% global coverage, but we’ll also begin developing live API connections with transport operators who offer this functionality. Combining those two approaches will allow us to offer the ideal user experience: truly global, door-to-door coverage; lightning fast results; and accurate fares and schedule information for users focused on specific travel dates.

Rome2rio’s Development Team Expands Again

In August this year we wrote about Miles Izzo joining our engineering team, and his experience and ability to tackle some of the pressing problems on our project list. Miles has already shipped several important improvements to the site, including the printable results feature, support for new transport types and performance improvements. We also mentioned we were on the lookout for two additional developers, and we’re pleased to say that search has been successful, with Marco Lui and Alan Lawrey joining our team this month.

Marco recently completed his PhD studies at Melbourne University where he published several papers on text modeling, analytics and classification while teaching classes and entering Kaggle competitions on the side. He is a long time fan of Rome2rio and told us “I always assumed that you were another Bay Area startup”. When he discovered that we were local, and hiring developers, he was quick to apply. Marco’s scientific and research background will be invaluable for many of the challenging data science and analytics related projects we’ve got going.

Alan completed a Computer Science degree at LaTrobe University in 2011 and has been working on back and front end web development at migenius, a Melbourne company that provides ray tracing web solutions. We love that he’s clearly passionate about coding and works on a variety of projects in his free time.

Marco and Alan

Alan and Marco outside the AKM building (home to inspire9 and Rome2rio)

We have been delighted by the success we’ve had over the last couple of months finding experienced, talented developers here in Melbourne. We have employed the same recruiting, interviewing and leveling processes that Bernie and I used at Microsoft and all our hires would certainly pass the bar for hiring at the Redmond tech giant.

Melbourne’s universities continue to produce top quality computer scientists. The absence of tech giants such as Google, Facebook and Twitter competing for local talent certainly makes it easier for Melbourne based tech start-ups to stand out as employers and set a high bar for recruiting. We are certainly not the only ones observing this – just yesterday Evernote announced it plans to open an office in Australia to “take advantage of what it considers to be an under-exploited developer and engineering talent pool”.