Fig 1.
Theoretical representation of satisfaction.
Fig 2.
Principle of satisfaction layers.
Fig 3.
Production cycle: Extraction → Transformation → Use (distribution—provision) → Re-use → End of Life (Destruction).
Table 1.
The P-Matrix (Predictors per cycle “moments”).
Fig 4.
Process life cycle.
Fig 5.
Flow chart for digital, economic, and ecological sobrieties.
Fig 6.
CNN architecture.
Fig 7.
Conversion tables from energy to cost [55] and from energy to CO2eq emission [56].
Fig 8.
Conversion table from energy to CO2eq emission.
Fig 9.
Moderation triangle through sobriety.
Fig 10.
The cost of training the algorithm is based on the type of energy source and the size of the training dataset.
Fig 11.
Cost difference and its rate between 37,440 and 49,984 images based on the type of energy source.
Fig 12.
Digital sobriety.
Fig 13.
CO2eq emission for training the algorithm based on the type of energy source and on the size of the dataset.
Fig 14.
Difference of CO2eq emission and its rate between 37,440 and 49,984 images, for training the algorithm, for various types of energy sources.