Difference between revisions of "Environment"

Revision as of 11:58, 6 August 2020

Open Food Facts is embarked on an effort to access the environmental impact of products. In summary, we need to

Categorize products in order to be able to use the lifecycle analysis averages that we already have.
Extract every little bit of information that we can from packagings, Including environmental labels and ingredients, both useful to have a better assessment of impact.
Improve our taxonomy system to be able to decipher it, and add explanations.
Gather lifecycle analysis data for countries outside of France or find ways to reverse-engineer lifecycle analysis files to tweak the parameters at very large scale

Adding food categories to assess the carbon footprint

In order to be able to apply the lifecycle analysis values, we need to know precisely the category of a product. You can help using various tools on Open Food Facts:

Adding food categories using the app

The official app for Android and iPhone shows category prediction when you are logged in.
You can answer them easily by tapping yes no or maybe.
You can even answer them if you don't have any products in your hands. For more instructions, see this page.

Adding food categories using Hunger Games

You can use Hunger Games to validate category predictions in mass.

Full list of OFF categories linked to an Agribalyse category

Adding food categories using the mass edit extension

You can use the Chrome extension to mass add categories. As this tool is powerful, its usage is restricted. Please contact @teolemon to get access.

Important note:Even if the product has a category, having a more granular category may be useful, as we don't have lifecycle analysis for overly broad categories.

Environmental labels

We need to collect environmental labels.

We have a logo annotation tool to help you collect labels at large scale.
You can also validate labels using Hunger Games.

Validation using Hunger Games logos

Please read the guidelines before starting: Guidelines

Random mode

EU Organic - Max Havelaar - AB Bio - Dolphin Safe - Canada Organic - Ohne Gentechnik - EG-Öko-Verordnung - Bleu Blanc Cœur - Le Porc Français - Label Rouge - Lapin de France - Produit en Bretagne - RSPO - Haute Valeur Environnementale - UTZ Certified - MSC

1-PET 2-HDPE 5-PP

Validation using Hunger Games labels

This is only a sub-selection of all the environnemental labels that might be relevant.

Header text	Header text	Header text	Header text
Organic	UTZ Certified	EU organic	Label Rouge
Haute Valeur Environnementale	Bleu Blanc Coeur	RSPO (palm oil)	Max Havelaar
Faitrade	Max Havelaar	Rainforest Alliance	MSC
AB Bio	French pork	EU agriculture - only EU origins	ASC

Categorize environmental labels in the labels taxonomy

It's now possible to add categories to labels to help the apps and website display it in a nicer way. Categorizing environmental labels will greatly help with that. Categorize environmental labels on GitHub

en:Label Rouge
ca:Etiqueta Vermella
de:Label Rouge
es:Label Rouge
fi:Label Rouge
fr:Label Rouge
nl:Label Rouge
wikidata:en:Q3214309
label_categories:en: en:Quality, en:Environment

Getting information and logos for environmental labels

Research the logo

- provide 1-line bullet points of what it means (there can be several bullet points)
- Find a icon idea for each bullet point (not the icon itself, but an idea for the icon) using The Noun Project
- Add the Wikipedia/Wikidata page to the label taxonomy if it's not already available

Get the image for the logo

- Get the logo in SVG (or PNG otherwise): If we don't have a SVG logo, we'd need one, if possible like this: label-name-as-it-is-in-the-openfoodfacts-url.123x90.svg The height comes in second and should always be 90. You need to calculate the ratio to find out the width. if the logo is 246x180, you need to divide the height by 2 to get to 90, so you need to put height 90, and width 246/2=123.
Send your research to @teolemon or commit it directly to GitHub if you are familiar with it.

Assessing product location

Processing locations & packaging codes

There are many traceability codes available on the food packaging. Those usually indicate the plant location which is helpful to compute distance the product has traveled. For instance, we do a map of all worldwide and European plants called Made Near Me. Some countries are still missing or not deployed.

TODO: Your help would be welcome (involves data harvesting and cleaning from official websites)

EMB codes and EU Codes
Madenear.me uses OFF data to indicate where the product was packaged

Origins of ingredients

Often times, the origin of ingredients is not in the ingredient list itself. It can be written on the packaging.

TODO: Create additional REGEX for Robotoff

Stores and brands

Those can be useful to get a coarse idea of location.

TODO: Improve the stores taxonomy worldwide and in France

LCA files produced by other institutions

We are lucky to get lifecycle analysis values for France but we need a similar values for other countries in the world. If you'd like to help with that, even for countries other than your own, please get in touch on Slack.

TODO: Get LCA values for various countries

Packaging

TODO:

We need to flesh out the packaging taxonomy that is currently on GitHub (TODO: Insert URL)

Fishing

See Fishing