Open car accident claims data from IF?

Thoughts and experiences from the Aaltoes Open Data Hackathon with IF

I got really excited last week when I noticed that IF, a Finnish insurance company, will open some of their car accident claims data in a hackathon. The topic was also interesting regarding some of our clients at Avaus, so I participated the event and among other things made a shinyapp for visualizing some of the claims data.

However, it turned out that the data could be used only during the hackathon, so it wasn’t quite open data yet. But Anders Stenbäck, the organizer from IF, said he is working hard to get the data set open. In his previous work at Sanoma he was involved in organizing the HS Open events where I also was participated a few times. So look forward to interesting data openings from IF in the near future!

Anyway, the hackathon, co-organized with Aalto Entrepreneurship Society and Tesla, was a really nice experience. The goal was to let the participants develop new services for car drivers to be used daily, focusing on connected cars, and the claims data was provided to inspire new ideas. The main prize was a trip to California and visit to Tesla’s HQ, not bad!

In the Friday evening’s warmup session we heard an interesting presentation from Rightware, a Finnish company developing next generation user interfaces for connected cars. IF also introduced their on board diagnostics sensor, which monitors car driving behaviour. A sample of this data was shown in the event.

I managed to miss the presentation of the winners, team Murikka, but hopefully the videos will be posted online soon. The second prize went to an app that enabled simple communication with adjacent cars. For instance when a nice person gives you way on the road, you could send her a smile and “thanks” via the app. With some social elements the app aims to encourage good behaviour in traffic, which is nice.

The third prize went to Ifify, an app that quites down the music in your car when you are speeding. Their very very impressive and cool video was quite an achievement in one weekend. The team has some background in using open data in cool apps, their Ihana Helsinki augmented reality app winning a prize in Apps4Finland in 2011.

In the end only maybe three teams actually used the claims data. One reason was that the data was quite limited, for instance location was given only on municipality level. Anders promised that they will try to get exact gps locations in a future data set, which would be cool and allow more accurate analysis of risk factors related to accidents.

So, no prizes for me. I was experimenting with a very data-driven approach to the hackathon, as that’s what I do best anyway. The core of my idea was to collect more data about accidents and, combined with other data sources, to use statistics to identify risky driving behaviour. Apps could then be used to encourage better behaviour, perhaps a’la Ifify. My approach and presentation stood out from the others and earned some nice compliments from the jury, but obviously was not enought to beat the best app ideas. I still learned a lot and already got a lot of ideas for improvement. Maybe we’ll see the next version in Apps4Finland this year…