due to GitHub storage constraints, we upload the majority of the data to DepositOnce and keep the more recent ride and profile data here on GitHub.
DepositOnce (06/2019-12/2020)
DepositOnce (01/2021-09/2021)
DepositOnce (10/2021-09/2022)
DepositOnce (10/2022-07/2023)
This project is part of the SimRa research project which includes the following subprojects:
- sirma-android: The SimRa app for Android.
- simra-ios: The SimRa app for iOS.
- backend: The SimRa backend software.
- dataset: Result data from the SimRa project.
- screenshots: Screenshots of both the iOS and Android app.
- SimRa-Visualization: Web application for visualizing the dataset.
In this project, we collect – with a strong focus on data protection and privacy – data on such near crashes to identify when and where bicyclists are especially at risk. We also aim to identify the main routes of bicycle traffic in Berlin. To obtain such data, we have developed a smartphone app that uses GPS information to track routes of bicyclists and the built-in acceleration sensors to pre-categorize near crashes. After their trip, users are asked to annotate and upload the collected data, pseudonymized per trip. For more information see our website.
The data from 06/2019 - 12/2020 can be found at https://doi.org/10.14279/depositonce-10605 The data from 01/2021 - 09/2021 can be found at https://doi.org/10.14279/depositonce-12452 The data from 10/2021 - 09/2022 can be found at https://doi.org/10.14279/depositonce-16439 Newer data is published in this repository.
The data collected in the SimRa project and published in this repository is made available under CC BY-NC 4.0 license (further information). While the license denies commercial use, we grant the following additional rights for commercial use: You are free to use and adapt the data for all journalistic purposes as if it were a non-commercial use.
All usage under the license and the extended rights described above are subject to the terms of use below. Any violation of the terms of use voids your right to use the data set.
-
For the purpose of the terms of use, ride data shall refer to the files that include the GPS trace and the incidents; profile data shall refer to the files that include demographic information such as gender and age group.
-
When you use the data you must not engage in any activities designed
- to identify movement profiles of individual app users,
- to correlate ride data with profile data, or
- to engage in other activities that might deanonymize cyclists or otherwise infringe our app users' privacy rights.
Should you discover any such information accidentally, you are required to delete that information immediately and to not make further use of the information. You are also required to assert that such information cannot be accessed by third parties before you delete it.
-
When you make this data set publicly accessible (either in original or adapted form) you must assert that your data copy also complies with these terms of use. For this, you can either use comparable terms of use or take technical measures such as data aggregation to assert the same outcome.
Each file has the information of app version number and file version number as it's first line. File version number indicates how often a ride has been uploaded or edited: #
The "Rides" folder contains the rides and the annotations of the incidents, that were detected by the app or added by the user. In a file, first the incident information are listed, followed by an empty line, followed by "=========================", followed by the the ride information.
incidents header: key,lat,lon,ts,bike,childCheckBox,trailerCheckBox,pLoc,incident,i1,i2,i3,i4,i5,i6,i7,i8,i9,scary,desc,i10
key: number of the incidents (starts at 0)
lat/lon: latitude and longitude (GPS location)
ts: timestamp (number of milliseconds from epoch)
bike: type of bicycle:
0 = not chosen
1 = City-/Trekking Bike
2 = Road Racing Bike
3 = E-Bike
4 = Recumbent Bicycle
5 = Freight Bicycle
6 = Tandem Bicycle
7 = Mountainbike
8 = Other
childCheckBox: 0, if no child is being transported on the bike, 1 otherwise.
trailerCheckBox: 0, if no trailer is attached at the bike,1 otherwise.
pLoc: Location of the phone during the ride:
0 = Pocket (default value)
1 = Handlebar
2 = Jacket pocket
3 = Hand
4 = Basket/Pannier
5 = Backpack/Bag
6 = Other
incident: Type of incident:
-5 = Dummy incident (if no incident is set, this is created to preserve bike, pLoc, childCheckBox and trailerCheckBox info)
0 = Nothing (default value)
1 = Close Pass
2 = Someone pulling in or out
3 = Near left or right hook
4 = Someone approaching head on
5 = Tailgating
6 = Near-Dooring
7 = Dodging an obstacle (e.g., a dog)
8 = Other (Please specify below)
i1-i10 are other participants involved in the incident. 1, if the according type of participant was involved, 0 otherwise.
i1 = Bus/Coach
i2 = Cyclist
i3 = Pedestrian
i4 = Delivery Van
i5 = Lorry/Truck
i6 = Motorcyclist
i7 = Car
i8 = Taxi/Cab
i9 = Other
i10 = Electric Scooter
scary: 1, if the incident was scary, 0 otherwise
desc: text description of the incident.
========================= (divides incident information and ride information)
Header: lat,lon,X,Y,Z,timeStamp,acc,a,b,c,obsDistanceLeft1,obsDistanceLeft2,obsDistanceRight1,obsDistanceRight2,obsClosePassEvent,XL,YL,ZL,RX,RY,RZ,RC
lat/lon: latitude and longitude (GPS location)
X/Y/Z: accelerometer sensor readings. Further information: https://developer.android.com/guide/topics/sensors/sensors_motion.html#sensors-raw-data
ts: timestamp (number of milliseconds from epoch)
acc: the radius of 68% confidence, meaning that there is a 68% chance that the true location is within that radius of the measured point (lat, lon) in meters
a/b/c: gyroscope sensor readings. Further information: https://developer.android.com/guide/topics/sensors/sensors_motion.html#sensors-motion-gyro
obsDistanceLeft1/obsDistanceLeft2/obsDistanceRight1/obsDistanceRight2: OpenBikeSensor readings on the left and ride sides in meter
obsClosePassEvent: Whether an obsClosePassEvent occured here
XL/YL/ZL: linear accelerometer values (X/Y/Z without gravity) Further information: https://developer.android.com/reference/android/hardware/SensorEvent#sensor.type_linear_acceleration
RX/RY/RZ/RC: rotation vector values. Further information: https://developer.android.com/reference/android/hardware/SensorEvent#sensor.type_rotation_vector
Profile: Contains demographical data of the user.
The header changed throughout the versions and can differ between iOS and Android Versions.
The following map shows which column names can appear at which index like this: =[column_name1, column_name2]:
0=[birth], 1=[gender], 2=[region], 3=[experience], 4=[numberOfRides], 5=[duration], 6=[numberOfIncidents], 7=[waitedTime, length], 8=[distance, idle], 9=[0, numberOfScary, co2, behaviour], 10=[0, 1, numberOfScary], 11=[0, 1, 2], 12=[1, 2, 3], 13=[2, 3, 4], 14=[3, 4, 5], 15=[4, 5, 6], 16=[5, 6, 7], 17=[6, 7, 8], 18=[7, 8, 9], 19=[8, 9, 10], 20=[11, 9, 10], 21=[11, 12, 10], 22=[11, 12, 13], 23=[12, 13, 14], 24=[13, 14, 15], 25=[14, 15, 16], 26=[15, 16, 17], 27=[16, 17, 18], 28=[17, 18, 19], 29=[18, 19, 20], 30=[19, 20, 21], 31=[22, 20, 21], 32=[22, 23, 21], 33=[22, 23], 34=[23, behaviour], 35=[numberOfScary] Example Header: birth,gender,region,experience,numberOfRides,duration,numberOfIncidents,waitedTime,distance,co2,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,behaviour,numberOfScary
birth: birth year group
0 = Please Choose
1 = after 2004
2 = 2000–2004
3 = 1995–1999
4 = 1990–1994
5 = 1985–1989
6 = 1980–1984
7 = 1975–1979
8 = 1970–1974
9 = 1965–1969
10 = 1960–1964
11 = 1955–1959
12 = 1950–1954
13 = before 1950
gender:
0 = Please Choose (default value)
1 = Male, 2 = Female, 3 = Other
region:
0 = Please Choose (default value)
1 = Berlin/Potsdam
2 = London
3 = Other
4 = Bern
5 = Pforzheim/Enzkreis
6 = Augsburg
7 = Ruhr Region
8 = Stuttgart
9 = Leipzig
10 = Wuppertal/Solingen/Remscheid
11 = Düsseldorf
12 = Eichwalde/Zeuthen/Schulzendorf
13 = Hannover
14 = Bielefeld
15 = Munich
16 = ZES Experimental
17 = Konstanz
18 = Weimar
experience: experience as a cyclist (in year groups):
0 = Please Choose (default value)
1 = More than 10 years
2 = 5 to 10 years
3 = 2 to 4 years
4 = Less than 2 years
numberOfRides: Number of rides a user has uploaded
duration: The duration of rides a user has uploaded (unit: milliseconds)
numberOfIncidents: The number of incidents a user has uploaded
waitedTime: The duration a user had to wait during his uploaded rides (unit: seconds)
distance: The distance of rides a user has uploaded (unit: meters)
co2: The co2 savings of rides a user has uploaded by not driving a car (unit: g)
0,1,...,23: The distribution of rides through the hours in a day. If a ride is from 2:34pm to 3:10pm, the values at 2 and 3 will be increased by 0.5
behaviour: How much the user follows the traffic rules. 0 - 5, where 0 is "never" and 5 is "always"
numberOfScary: The number of uploaded incidents that were labeled as scary by the user