Serving and visualizing LCA results on a web page

[romainsacchi]

The deliverable for this working group would be to have a working web-based application that can serve LCA results to users.

Specifically, there would be a landing page where:

• the user could pick one or several products/services,
• define a functional unit (reference flow, allocation method, impact assessment method, etc.),
• send this calculation setup to the LCA calculation app,
• and receive the results, which could be visualized using graphs
• and be exported in a variety of format (csv, ecospold, etc.).

The results would be cached for ulterior requests.

A nice-to-have feature would be to let the user submit data to be added to the triplestore database, either by manually entering the data or by a drag-and-drop of a csv file (in a pre-defined format first, but maybe using a ML text classifier in the future).

Foreseeable tasks:

• create a basic Flask project
• host the code on Github, with instructions for "reproducibility"
• set on a structure for the Flask project to reflect the future needed modules
• write basic user functions on the landing page (picking a product, defining a functional unit, etc.)
• make the LCA calculation module work with dummy matrices, or Exiobase matrix
• cache the matrix on first request
• register users' requests and results in a database
• install the Flask app on a droplet with Ubuntu or Fedora
• develop the converter module, effectively converting a triplestore extract into a 2D matrix
• make a simple widget to be embedded in third-party pages