Trafi, or Finnish Transport Safety Agency develops the safety of the transport system, promotes environmentally friendly transport solutions and is responsible for transport system regulatory duties. For example, in 2015 there were more than 2.6 million passenger cars in circulation in Finland. Trafi
- issues permits, regulations, approvals and decisions and prepares legal rules regarding the transport sector;
- arranges examinations, handles transport sector taxation and registration, and provides reliable information services;
- oversees the transport market as well as compliance with rules and regulations governing the transport system;
- participates in international co-operation;
- ensures the functionality of the transport system even in emergency conditions and when normal operations are disrupted;
- creates opportunities for innovative development of intelligent transport;
- informs the public of transport-related choices.
Trafi has begun to make various data content available for public use free of charge. In this case, the dataset “Open data for vehicles 4.7” covers details on vehicle registration, approval and also on technical data for all vehicles in road traffic. The dataset is licenced after CC BY 4.0 licence.
The business case was to utilize the Trafi dataset and provide suggestions, which would help setting up a new car dealership in greater Seinäjoki area in Finland. The area includes six municipalities (Kauhava, Lapua, Seinäjoki, Kuortane, Alavus and Ilmajoki) and the population is approximately 130 000.
Data-analysis was performed with Tableau and after loading the +800-megabyte dataset, it was filtered to include only years 2000-2015 and the municipalities mentioned earlier. However, Trafi uses numeric codes to identify municipalities. Therefore, it was necessary to lookup respected codes for each municipality. For years, the registration date was used as a filter.
The figure above shows the development of vehicle registrations by vehicle type from year 2000 to 2015. Blue colour is a normal car, orange is a light trailer and green is a van. Based on this information and to limit the prospected business scope of the car dealership, vehicles were filtered to include only types M1 and M1G (passenger cars / SUVs, max. 8 passengers + driver). The number of registered vehicles per year dropped significantly after the 2008 financial crisis and has failed to recover.
After this, the fuel type of a vehicle was inspected. Orange is diesel and blue is petrol.
As can be seen, popularity of diesel vehicles was increasing until 2008, when more than half of newly registered cars were using diesel. However, in the following years the share of petrol cars has dropped, while petrol cars have been more popular, taking more than two thirds of registered cars in 2015.
Together with the price development index of petrol and diesel, some additional insights can be gathered: since 2010, the price of diesel has increased compared to petrol. In addition to this, in Finland all diesel powered vehicles has to pay an additional tax. While new petrol engines have become more fuel efficient, he yearly kilometre limit when diesel becomes cheaper to drive than petrol has moved upwards.
When looking at the body types (wagon = red, sedan = orange, hatchback = green, purple = multi-purpose), the share of sedans has dropped from 40% to 15, while multi-purpose vehicles have increased from almost zero to 30%. Hatchbacks and wagons have remained popular as well.
Average co2 emissions have also decreased in both fuel types.
From 2010 to 2015 most popular car brands include everything from Audi to Volkswagen. It is interesting to notice, that despite the downward trend the brands have performed relatively similarly year to year.
However, when comparing cars from premium segment to the other group, diesel appears to the popular choice.
One of the biggest challenges with this data set was the quality of data. For example, Volkswagen was presented as “Volkswagen”, “volkswagen”, “VOLKSWAGEN” and “Volkswagen, VW”. As a result, the registration numbers per year varied based on the way Volkswagen was written in to the system. I solved this by grouping all possible Volkswagen-variants together under one variable. Similar issues are probable present with other car brands and model names.