Digital Decision Support for Refurbishment (2DSR)

Purpose and goal:

Refurbishment represents a key industrial pathway in the Circular Economy, but the decision to pursue circular pathways can be costly. This project is focused on data-driven development and secure exchange of data between process and actors. Drawing from data from a diverse consortium of actors in the furniture industry, we will use machine learning to develop an image-based decision support system (DSS) that expedites cost and environmental impact estimations that traditionally require costly expert assessment and/or extensive resources.

Expected results and effects:

This project will help Swedish furniture firms gain a competitive edge in the circular economy. The resulting decision support system will validate refurbishment decisions by quantitively estimating relative circularity and climate impact; it will addresses knowledge gaps about the relationship between circularity and environmental impact by applying a circularity metric and lifecycle assessment independently, but in parallel; and contribute to Agenda 2030, specifically goals 9, 12, and 13.

Approach and implementation:

The project involves several parallel work packages. Work Package 1 is responsible for project leadership and communication. Work Package 2 will develop a conceptual map for data collection and plans for data security.

Work Package 3 uses machine learning methods to observe patterns across various forms of image and text data to classify furniture articles, estimate cost, and estimate environmental impact. Work Package 4 uses lifecycle assessment to estimate the relative carbon dioxide impact of refurbishment decisions as compared to "buying new" furniture.