Android application that lets you track raw data for bike path quality analysis
The aim of this app is to track phone sensor data while riding a bike. These sensor data include
- geolocation including speed
- accelerometer
After you performed a bike activity (aka bike trip) you can review the bike activity and the tracked samples, specify the surface type and smoothness type of the bike path and upload the data. The uploaded raw data will be used to train a neural net which will be able to determine the surface type of a bike path as well as to detect anomalies such as potholes based on sensor data.
The app will only work with Android devices with Android 8.0 (API level 26) or above. On your Android device opt-in for installing unknown apps. There are two ways to install the app - from the bundled releases and from your local IDE.
- Download the latest
app-release.apkfrom the releases section - Open the file and follow the instructions
- Download and install the latest version of Android Studio
- Clone the repo
git clone git@github.com:fom-big-data-bike-path-quality/fom-big-data-bike-path-quality-android.git-
Open the project in Android Studio
-
Setup Firebase for Android as described here
-
Setup Mapbox Maps SDK as described here
-
Setup app signing as described here
-
Connect an Android device via USB cable
-
Run the app
On first start-up of the app you will be prompted to give permission to track location and activity transition (this is used for automatic tracking mode). Click CONTINUE in both cases and allow both while using the app. After that you will see the main screen. From the main screen other screens like the sensors screen and the log screen can be reached by swiping left or right. The settings screen is displayed after clicking the settings icon in the menu.
 |
 |
 |
|---|---|---|
| Permission rationale location | Permission rationale activity transition | Main screen |
Both sensors screen and log screen can help you understand measured values and actions performed by the app. On the sensors screen currently measured values are displayed. The log screen displays logging information about the app status. In the settings screen you can define preferred phone position and bike type. Furthermore you can decide whether tracking shall be started and stopped automatically. Also measurement intervals and sample sizes can be specified. The app can track biking data in two modes
- automatic tracking experimental will automatically start and stop bike activity tracking when you start or stop riding a bike
- manual tracking allows you to start and stop a bike activity whenever you want
 |
 |
 |
|---|---|---|
| Sensors screen | Log screen | Settings screen |
On the main screen you can can start manual tracking by clicking the green arrow button.
Important: in order to train the model pure data is needed. Try to ride on one surface type only when tracking a bike activity. Stop and restart the activity tracking when you switch the surface type. (Later you can set the surface type for each sample individually which might become cumbersome if you need to do it for too many samples)
In order to ensure a safe right without distractions you can switch to head-up mode by clicking on the square icon in the menu. In head-up mode your current speed, the current bumpiness of the surface and the collected samples are displayed. The color of the screen indicates how bumpy the surface currently is. You can return from head-up mode and start or stop tracking by using the buttons on the bottom.
 |
 |
 |
|---|---|---|
| HUD screen with smooth surface | HUD screen with intermediate surface | HUD screen with rough surface |
Once you completed the tracking of a bike activity you can see the details by clicking on its list entry on the main screen. From here you can specify the surface type and the smoothness type of the bike activity. Phone position and bike type are prefilled with the values you specified in the settings screen.
 |
 |
 |
|---|---|---|
| Activity details | Surface type selection | Smoothness type selection |
The surface type for the bike activity you can overridden for single parts. You can do so by selecting a sample from the list and click the dashed button next to it. After setting all fields you can upload your tracking results by hitting the green floating action button.
 |
 |
|---|---|
| Activity details with focus | Activity upload |
See the open issues for a list of proposed features (and known issues).
- 1.13.0 Material You support
- 1.12.0 Phone position and bike type preferences
- 1.11.0 Head-up mode
- 1.10.0 Bike activity sample span highlight
- 1.9.0 Bike activity sample color code
- 1.8.0 Invalid bike activity sample indicator
- 1.7.0 Under-lab-conditions flag
- 1.6.0 Tracking fixes
- 1.5.0 Firebase upload
- 1.4.0 Phone position and bike type selection
- 1.3.0 Configurable tracking intervals
- 1.2.0 Activity sample selection
- 1.1.0 Anonymous user identification
- 1.0.0 Basic functionality
- Activity transition detection
- Accelerometer sensor measurement
- Location sensor measurement
- Bike activity tracking - in background, manual and automatic
- Bike activity samples on a map
- Surface type and smoothness type selection
- app version for the broad public that does not require manual input
Since this project is part of an ongoing Master's thesis contributions are not possible as for now.
Distributed under the GPLv3 License. See LICENSE.md for more information.
Florian Schwanz - florian.schwanz@gmail.com
Icon made by Freepik from www.flaticon.com