-
Bike Monitor
Android App that lets you record your bike journey
-
+GPS
Map your bike trip
-
Stress map
Find which streets you were stressed
The number of bicycles is increasing in many cities, making them cleaner and more environmental friendly.
But this increase of bike usage has also lead to an increase of bicycle related accidents in Portugal
To accommodate this increase, cities need to adapt their bike related infrastructures. One way to identify the streets that need intervention, is trough a study of cyclist risk perception
Number of bicycle accidents in Portugal:
1914 in 2014
1762 in 2013
1487 in 2012
1485 in 2011
1227 in 2010
in: ASNR
In order to study the urban ciclysts we mounted a system composed by three components.
Smartphones are widely
available and equipped with several sensors!
From the smartphone we collect:
-GPS Coordinates
-Acceleration on 3 axis (Accelerometer)
-Rotation (Gyroscope)
-Image (Camera)
-Date
With the captured images from the action we can evaluate the ciclyst surroundings
With the chest strap we collect the cyclist ECG that can inform us about the cyclist stress
The images are captured with the action cam mounted on the bike frame
Using Lukas-Kanade algorithm, the optical flow of the captured images is calculated
A descriptor was built based on the size and direction of the flow vectors
Based on the image we can know if the cyclist:
-is stopped
-is turning right
-is turning left
-is going forward with no obstacles
-is being overcome by its left side
-is passing through ground irregularities
ECG is recorded using the heart monitoring chest strap
The R-R Intervals, interbeat intervals, are then extracted from the recorded ECG
The power spectral density is then calculated using the RR values
The stress of the cyclist is then calculated using the HRV, Heart Rate Variability
During this work several test trips were performed, with the final prototype a database was created.
To get the database, send us an email.
2
10
+120
+40 km
List of publications during the course of the Smartbike Project.
Costa, M., Quintino Ferreira, B. and Marques, M., “A Context Aware and Video-Based Risk Descriptor for Cyclists,” in 2017 IEEE 20th International Conference on Intelligent Transportation, Yokohama, Japan, October 2017
Costa, M., “Video-Based Risk Assessment for Cyclists,” M.S. thesis, Dept. Elect. And Comp. Eng., UTL, IST, Lisbon, 2017
Vieira, P., Costeira, J.P., Brandão, S. and Marques, M., "SMARTcycling: Assessing cyclists' driving experience”. In Intelligent Vehicles Symposium (IV), 2016 IEEE (pp. 1321-1326), June 2016
P. M. S. Vieira, “Percepção do risco em ambiente rodoviário urbano,” M.S thesis, Dept. Elect. Eng., UTL, IST, Lisbon, 2015
* Required
Instituto Superior Técnico
ISR
Portugal
smartbikeisr@gmail.com
