«routeRANK works like a chess computer»

routeRANK founder and chairman Jochen Mundinger talks about his idea for routeRANK, how CO2 emissions are calculated and how his mathematical background helps him in his work.

Jochen, how was routeRANK born?

When I was working at the EPFL I had to book a conference journey to Poland. I did not know the place, and I spent half an afternoon researching various travel options. In the end I still was not sure whether I had found the best solution. Even after booking the journey I checked other options, but have still found nothing.

As a mathematician you then sit down with pen and paper and start looking for ideas. I consulted both technical and business specialists and the reactions were positive. Somehow, I slipped in bit by bit.

So the idea emerged from a personal need. When did you decide to start a business?

That is not an overnight decision. Several weeks went by in which I was looking for existing solutions and was asking myself how to tackle the problem algorithmically. As I said, I consulted specialists, friends and family. Step by step, everything started coming together and my experiences were positive throughout, so I invested more and more time.

What is the challenge in developing routeRANK’s algorithms? Why cannot anybody simply recreate it?

There is a huge number of possibilities to travel from A to B; even for quite simple routes you can find thousands of ways. One cannot just calculate all of them. In the end we filed two patents. But how we do it exactly remains our secret.

We would like to know more precisely.

The algorithm focuses on the «interesting» routes and drops the uninteresting ones. In a way, it is similar to a chess computer, in that it does not calculate all the possible moves, but only the promising ones.

Where does the data come from?

Depending on the application – for example www.routerank.com or WWF Travel Helper – we use different data in different combinations. It also depends on whether you travel by car, public transportation or plane. For car data, for example, we use Via Michelin.

And how do you calculate CO2 emissions?

After looking closely at the state-of-the-art of research in this field, we decided on a model developed by the IFEU Heidelberg, which we adapted to our needs. We extended this model with data from the European Commission, from NPOs, transport companies and universities accross Europe. Independent researchers have examined our methodology and approved it.

What is more important, the algorithm or the aggregation?

Both are important and required. The algorithm on its own is a great, but if it does not have any data to work on, it cannot provide interesting results. On the other hand, if you have only data but no aggregation and optimization, that does not help either. In fact, the user already has the means to find all the information s/he needs online, but might have to spend four hours with the computer to find the best route. This is where routeRANK can help.

So far routeRANK only connects individual cities, stations and airports. Have you thought about door-to-door connections?

We could already offer this now, but it is a question of costs concerning geodata. However, we offer door-to-door connections for our customized versions, if a client asks for it.

How long did the development of routeRANK take?

It has been three years now. I hired people early on and soon there were five of us.

That is an enormous investment.

It is very complex. There is the optimization. Then we have to handle a variety of data from different sources, on top of that the web application with database and server, and then the challenge of keeping the user interface simple.

Are requests live or do you store certain data on your servers? The roads from Paris to Berlin, for instance, don’t change every day.

It depends. Regarding road traffic, for example, it depends on whether the current traffic situation is considered. At the moment, routeRANK does not do this. But if we implement this feature it requires live requests.

Is routeRANK finished or still developing?

There is still a lot to do. For example the customized B2B versions we offer to companies and organizations. Even apart from this I can think of some features I would like to have on our website. We listen to our users to find out what is most relevant.

But your partners have a rather green angle?

The software as such does not make a difference between ranking your routes by CO2, by price or by time. With routeRANK, users can save all of money, time and emissions. For me personally CO2 was an integral part from the beginning, though. And we have accomplished a lot in the green sector in particular, so I am happy about that.

Where are the differences between routeRANK’s B2B and B2C versions?

B2B are customized versions for companies and organizations, and can differ in branding as well as data and features.

Do you not jeopardize your B2B business by offering a public website? Why should anybody purchase a customized version if they can just access the public site?

The versions are customized. For example, one customer is interested in optimizing work time. While driving one cannot work at all; when traveling by plane, maybe one can work for one and a half hours. And when traveling by train one can work four hours, say. This is an example of a customized feature.

You are a mathematician. Does that help you in your day-to-day work? To you, is routeRANK an optimization problem?

Yes, one might say so. I do think my mathematical background has an influence on my everyday life.

If you compare your experience of having to book a journey to Poland with the service you offer today: Do you think you have realized your vision?

Yes. We might not have integrated every location and every connection yet, but overall I am very happy with what we have accomplished so far.

If you enjoyed this post, please consider to leave a comment or subscribe to the feed and get future articles delivered to your feed reader.