CSE 6242 Final Project, FALL 2017

TrailBlazer: a web based application to help bikers, hikers and runners to find the most exciting tracks to train!

Developped by Alexandre PALO, Alex MUELLER, Guillaume BROGGI & Tianyi ZHENG

DESCRIPTION

TrailBlazer is a project which aims to help bikers, hikers and runners to find the best tracks to train outside. Imagine you are new to an area, how to find where to practice your favorite sport? The application will make the most of existing sport tracks and points of interest in our database and will compute the best track regarding the preferences you will give.

INSTALLATION

Front end application

Installation

Preleminaries

Be sure to have NodeJs and Yarn installed (or npm).

Packages installation

Install serve package

• yarn global add serve (or npm install -g serve). If needed, run this command with administrator privileges or sudo.

Back end application

Installation

Preleminaries

Be sure to have Python 2.7 installed and pip installer for Python 2.7 too. In this tutorial, we assume that the command python launches Python 2.7 and that pip installs for Python 2.7 only.

Virtual env

The back end for our application is going to be installed in an python virtual environment to ensure that every package is here with the right version.

Install virtualenv globally:

• Windows: pip install virtualenv within an administrator cmd
• MacOs & Linux based: sudo su & pip install virtualenv

Go to the folder of the project, and the backend subfolder:

• cd PATH_TO_FOLDER_DOWNLOADED_UNZIPPED/backEnd

Create a new virtual environment and activate it:

• mkdir virtualenv
• virtualenv virtualenv/backEnd
• cd virtualenv/backEnd/bin
• source activate
• cd ../../..

Install all packages needed for the back end of our application, still in the virtualenv:

• Make sure that python --versionreturns you Python 2.7.X, X is 10 or higher.
• pip install django==1.11 geopy mongoengine geohash networkx
• pip install djangorestframework django-rest-framework-mongoengine
• pip install django-cors-headers

Leave the virtual environment:

• deactivate

Launch application

To launch the application, two consoles should be opened (or to tabs). These two consoles must be up and running the whole time you use the application.

First console, front end application

• cd PATH_TO_FOLDER_DOWNLOADED_UNZIPPED/frontEnd && serve -s build

Second console, back end application

• cd PATH_TO_FOLDER_DOWNLOADED_UNZIPPED/backEnd/virtualenv/backEnd/bin && source activate
• cd ../../.. && python manage.py runserver

EXECUTION

Go to the following url within your favorite browser (please prefer Chrome or Firefox): http://localhost:5000

Introduction

On order to simplify the preprocessing system, we have limited the application to a certain area of the world: an area around Chambéry in FRANCE.

This area is a good solution: the number of POI and Tracks gathered in this place is very important due to the geographic situation of this area. It's a famous place in FRANCE to practice mountain bike, in the French Alps.

Request tracks

• On the first splash screen, click the button "LET'S GET STARTED!".

• The "Settings" form let you input your preferences to find the best track for you. In order to do so, the only two mandatory information are the begin location and the distance to compute results.

• To entire the desire begin location, two solutions are possible: you can either try to enter an address in the field and select the best one using the autocompletion system (it's case sensitive), or you can click on the small "location" icon at the right of the field, click a point on the map and select "SET" to choose this location as begin location. As said, the application is only currently usable and viable around Chambéry, thus please use the second solution and click near the city of Chambéry. The map is centered on this place by default.

• Moreover, to compute best results you can include "POI" and "Tracks" from our database and select a weight for them. POI are points of interest that are interesting to be visited. Tracks are existing mountain bike tracks that help the algorithm to decide. Both of them should be include, that's the aim of our project to make the most of existing information.

• When everything is set up, click the button "SUBMIT". The application is going to ask your algorithm to find the best track regarding information you have chosen. Waiting until the loading screen stopped. This can take some time, up to 2-3 minutes for a 20 km request.

• When the algorithm has responded, you can see proposed solutions in the field "Track selector". For each track, you can click on it to activate or deactivate the solution. If the track is activated, it will appear on the map. Moreover, you can click the target icon at the right of each track to move the map to the beginning of the track. You can browse the solution with your mouse over the graph or directly over the drawing on the map.

• When you've selected an interesting solution, you can click the button "DOWNLOAD GPX" to download a standard gpx file and use it within your gps device (including your phone) and go outside to train!

• Be careful and indulgent, our application is only a beta version restricted to a small area. At any time, if an error occurs, juste reload the page (F5) in your browser. Moreover, if the backend seems to be blocked, go into the right console and restart the command: CTRL+C or CMD+C to stop the server and python manage.py runserver in the virtual environment to relaunch. Note that OpenStreetMap data are only around Chambéry, even if UtagawaVtt and Geocaching was scrapped in whole France.

• Have fun!