A fun weekend project by Emelie Hofland and Jaime González-Arintero. Created during the DB Open Data Hackathon in Berlin, on May 12-13, 2017.
Sometimes, you want to enjoy a quiet train trip, get some work done, read something, or chill. Or just have a siesta!
But suddenly, a bunch of noisy (and annoying) people jumps in the train, and your peace is gone. Oh, that weekly football match that attracts all hooligans. And them all want to sit close to you. Say "bye, bye" to that peaceful moment, my friend.
Wouldn't it be nice to know in advance what to expect from your trip?
The goal of this project is to collect data from a noise / loudness sensor, and combine it with other (travel) data (weekday, time, segment, and more). Eventually, predictions can be made on how busy or quiet your next trip will be.
For this, real data from the Deutsche Bahn Fahrplan API was used, as well as an affordable loudness sensor (to have a "real world" proof of concept).
All the data was later processed and represented with R. The sensor node was programmed in Arduino. An additional dashboard for real-time notifications was built in Node-RED, and the "predictive" dashboard (with all the routes) was created using Microsoft Power BI.
The following modules are represented on the architectural diagram:
- Loudness sensor to acquire real-time noise values from the train. A WiFi microcontroller takes care of sending such data to the cloud using the MQTT protocol.
- Speaker that plays a "please be quiet" message when the noise level on the train is too high.
- Script that converts the information about the most up-to-date train lines from the Deutsche Bahn Fahrplan API into GTFS.
Additionally, these optional modules may be implemented, but are not represented on this diagram:
- Data selection and representation dashboard, built in Microsoft Power BI.
- Real-time notifications and noise level dashboard, built in Node-RED.
The presentation is available in Google Slides (Chrome is recommended; other browsers such as Safari may show format issues).
A static PDF version can be downloaded here.
The additional pictures used in this presentation are available on this Google Drive folder, in the pics directory.
To inform the passengers that the loudness levels are too high, and kindly remind them to remain quiet, a voice notification will be played. Luckily, there are many TTS (Text To Speech) services online out there, including some free ones.
In this project, the service texttospeech was used in order to generate an audio file we could use. The transcription of such message can be found below:
-
English version: Dear Passengers, I would like to kindly remind you that this is a silent wagon.
-
German version: Liebe Passagiere, ich möchte Sie freundlich daran erinnern, dass Sie sich in einem stillen Wagen befinden. Vielen dank für Ihre verständnis.
For this purpose, the module mixer from pygame has been used in a Python script. The following code snipped loads such library, as well as the MP3 file, and it plays the file.
# Load the required library
from pygame import mixer
mixer.init()
mixer.music.load('<PATH_TO_MP3_FILE>')
mixer.music.play()Instead of <PATH_TO_MP3_FILE>, introduce the path to the chosen MP3 file (remember to add the file extension). For example, when this project was presented the script was executed in a Windows machine, and thus, the path was: D:/railax-take-it-easy/assets/silent-coupe-deutsch.mp3
IMPORTANT: Don't forget to run mixer.music.load('<PATH_TO_MP3_FILE>') each time before playing the file. Otherwise, the playback will not work.
This excerpt will later be integrated in the complete script that reads the values from the loudness sensor, and reacts accordingly. But first we need to connect the sensor itself.
---DOCS IN PROGRESS!---
To connect the noise / loudness sensor to the cloud, and start sendind real-time values, head to this tutorial.
---DOCS IN PROGRESS!---
We collected some live stream sound values from our sensor in R and explored the values.
It seems that the value does not often go over 600 (the original value we had in mind to set as a trigger), even in a noisy room. Therefore we put the trigger now at 400.
---DOCS IN PROGRESS!---
In the noise_mp3_reaction_python.py you can find the script that makes the mp3 go off every time the sound sensor hit a value over 400 more than three times.
---DOCS IN PROGRESS!---
To make it easy to work with the data in R via the gtfsr library - we converted the Fahrplan API data for this month into GTFS data - with the help of this python script.
- Add how-to for the sensor node.
- Add offline testing sketches as well.
- Add how-to for the Node-RED dashboard.
- Add link to the GTFS wiki page on the Solution Architecture section.
- Add links to the DB Fahrplan API.
- Check the proper stylized name of texttospeech.
Copyright (C) 2017 Emelie Hofland [email protected], Jaime González-Arintero [email protected]
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
Except as contained in this notice, the name(s) of the above copyright holders shall not be used in advertising or otherwise to promote the sale, use or other dealings in this Software without prior written authorization.
THE SOFTWARE IS PROVIDED "AS IS," WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.