Some days ago I had an opportunity to have a chat with Christian and Iulian, the two members of the team KCL from Kings College in London who created a system called Pioneer which aims to reduce pollution in large cities by reducing traffic congestion.

You can find the video they submitted for the Imagine Cup final (and also vote for their entry if you like) here:


I really like their work and their project, and wanted to publish some of the questions I asked them.

Q: What is the problem you’re solving and why did you choose it?

The problem we tackled is air pollution in large cities. As we are students living in London, we are directly confronted with this problem. This is why we saw how urgent it is.

Q: That sounds like a tough task, so how exactly did you solve it?

Well, first of all, we had to find out why big cities are so polluted. We found statistics from Greenpeace saying that more than 80% of urban pollution is caused by cars. And in particular, the pollution comes from traffic congestion.
Our solution aims to minimize traffic congestion through a central car routing system. By minimizing traffic congestion we could decrease air pollution significantly!

Q: That sounds interesting, so how exactly would such a routing system look like?

Basically, you have to think of it like a navigation system in your car. From a users perspective it looks the same way as other navigation systems look like. But on conventional navigation systems, routes are calculated on your small device, where our solution is based on a server that knows the position of all cars at any time. Based on that information, the server calculates your optimal route and sends it to your device.

Q: So, on the bottom line, the user will get a route that avoids heavy traffic and therefore he saves time, fuel and the environment?

That’s right, but that’s not everything: After we did first simulations, even we were surprised how efficient pioneer is, so we went even further by taking traffic lights into account. Pioneer is now a platform for the future city. It is able to switch traffic lights according to real time traffic data, which decreases congestion even more.

Q: That sounds brilliant, but can you give any numbers? For instance, by how much do you reduce traffic congestion?

Yes, we implemented a simulation module that gives some numbers. The average person drives the car 76 minutes every day. Our simulations show that with pioneer, everyone can save up to 10% of this time.

Q: And do you have any plans with pioneer for the future?

Oh yes, indeed we have big plans with pioneer. The next step is what we call the “Dynamic City”.

Q: What does that mean?

Let me explain it with an example: Imagine you are driving to work in the morning rush hour and you’re taking one of the major roads into the city centre. Your road has 5 lanes, two of them going into the city and 3 out of the city. Our system would change the direction of lanes according to the traffic needs. In this situation, it could would happen that the system automatically changes the road so that 4 lanes are going into the city. That would dramatically decrease congestion on that road.

Q: This is all good an well but… aren’t there already too many GPS navigation systems?

Absolutely. This is why our system is based on open APIs that 3rd party can use to leverage it. Companies like TomTom could tap into it to get routing information in real time and participate to reducing congestion.

Also, how many times did it happen to you to get in your car and not remembering the exact address you are going to, or not finding it on the map? The pioneer’s architecture allows to do some very nice things such as planning a route while sitting at your PC and send it to the mobile or GPS device making it easier to plan a trip without bad surprises.

Q: I know that planning routes taking in account real time traffic information is not a trivial problem.

Yes, and this is probably the biggest strength of our solution. 

Conventional algorithms are far too slow for large numbers of queries on large graphs like the ones representing large cities.

Our solution have a pre-processing step which makes it efficient enough to serve hundreds of queries a second on even very large city graphs which allows us to make it possible to do real time routing so that we can include traffic data in the calculations.

The research paper describing the details of how this has been achieved will be published early next year.

Q: Wow, I think this idea has potential in the future. I wish you all the best and good luck! Before we close though, can you tell me how and why you decided to participate to the Imagine Cup?

We heard about the Imagine Cup last year and this year we decided to submit an entry because we though we had a good idea and a great technology. We also think it is a great opportunity for us to make a good experience on a real project which we could potentially turn in a real product and business. Last but not least we think it is a great opportunity for us to have something in our resume that will show our participation and commitment and that will surely be very interesting when we finish our studies and start looking for a job.

To learn more, go to the project website where the team will publish info about their work