Mobility patterns analysis and prediction service development started in
European project Mobis (EU FP7, GA No 318452) and then continued in Optimum
(EU H2020, GA No 636160-2). The aim
of both projects was to persuade users towards sustainable mobility.
The service provides analysis of visited locations and routes on individual, group and global (all users) level. It can also make next location prediction for individual user based on historical data. Real-time data from Optimum users is also used for traffic event detection (traffic jams).
Collected data is used exclusively for research purposes.
We collect data from several sources. One of the sources are mobile phones with installed Mobility patterns or Optimum app. There, we collect the data in two ways:
Automatically collected data is used for analysis of mobility patterns.
Records containing GPS location, activity data and sensor (accelerometer
and magnetometer) samples are passed to our clustering engine where
the records are clustered into staypoints and paths.
Sensor samples are used to infer your transportation mode, e.g. to
determine whether you are travelling by car, by bus or by train.
Names of your personal locations are only used for display in app or web GUI to provide you with first-class personal experience. Modalities of your routes are used for display and as labels for further training of our classification models. For training we only use sensor samples and modality labels, which cannot be linked back to you.
If you don't want to participate in the project any more, you can simply
uninstall the collection app (Optimum, Mobility Patterns etc.)
If you want to delete personal data that was collected from you, you can do so from the main page in the top-right menu. This will delete the raw data as well as the processed data. Note that this action is irreversible. Note, that some of the data (e.g. accelerometer samples with labels), which is completely anonymous and cannot be linked back to you, is not deleted.
If you want to delete your account together with the personal data that was collected from you, you can again select the option in the main page from the top-right menu. This action, too is irreversible.
You can also download the data that was collected from your mobile phone. The data is in JSON format.
NextPin is developed at Jozef Stefan Institute, Artifical Intelligence laboratory.The following people contributed substantially to the project (in alphabetical order): Jan Berčič, Luka Bradeško, Sebastjan Fabijan, Zala Herga, Blaž Kažič, Matej Senožetnik, Tine Šubic and Jasna Urbančič.