This dataset was initiated in 2019 and is not maintained anymore. Its goal was to introduce one of the first apps for Amazon Alexa in Luxembourg. This project aimed at delivering a real use-case of local services on a voice assistant platform, focusing on waste pickup calendars. The dataset does not cover the entire country, as several other websites/sources should be crawled and consolidated to have a complete picture.

Code source and docs are available on GitHub : https://github.com/Juxz/luxwastecalendar

Since then, many municipalities have digitised their waste pickup calendars and offer great mobile services.

The first challenge of the project was accessing the raw data; at that time, the only choice was to scrap it from official websites. So I developed a nodejs modular scraping tool that connects to multiple sources:

1. HTML from sidec.lu using cheerio library
2. json from valorlux.lu
3. ICS files from vdl.lu using node-ical library

When scraping is complete, the tool unifies all results into a single format, normalizes pickup types, matches against the CACLR address database and writes 1 json file per postal code in a simple format :

[
  {
    "uid": "5e8a5f0732fc6",
    "event_date": "1608073200000",
    "city": "Luxembourg",
    "location": "Côte d'Eich",
    "streetNumbers": "1-25, 2-24",
    "codepostal": 1450,
    "summary": "BULKY"
  },
  {
    "uid": "5e8a5f074f2c3",
    "event_date": "1608505200000",
    "city": "Luxembourg",
    "location": "Côte d'Eich",
    "streetNumbers": "1-25, 2-24",
    "codepostal": 1450,
    "summary": "PAPER"
  }
]