At Pickler we understand that collecting data is a time consuming job. Therefore, we help users to fill in quickly the gaps in the missing data, so that they can improve it along the way when there is better data at hand.
Examples on our current predictive algorithms;
Predicting the material from the background database. Since we add common tags and brand names to materials of the background database, it gets easier to match to the clients data (E.g. Polyethylene terephthalate -> PET).
Predicting material shares based on the packaging component (E.g. lid = 3%)
Predicting production energy usage based on packaging category and material type. (E.g. PET and Bottle, result in prediction on Extrusion Blow Molding as production process)
Predicting transport method and distances based on the location of the production facility. (E.g. X point in China results in the transport distance by truck to the nearest container port, distance by container ship to nearest port of Pickler’s client warehouse, distance from port with truck to warehouse Pikcler’s client warehouse)
Predicting end of life; for each raw material we predict what the most common used end-of-life scenario is in Europe. (E.g. Plastics are burned with electricity recovery, and we count recycling credits on the input side)
On the roadmap are more predictive algorithms to help to fill in the data easily. E.g.;
Predicting the most used material types per (packaging) category
Predict material shares based on similar products
Predict production methods based on similar products
However, the user (packaging supplier) is still responsible for that the input data, such as production location or transport method, is correct, and has to be able to back this up with proof.