Digital Decision Support for Refurbishment (2DSR)
| Reference number | |
| Coordinator | RISE Research Institutes of Sweden AB - RISE |
| Funding from Vinnova | SEK 3 980 849 |
| Project duration | September 2022 - March 2026 |
| Status | Completed |
| Venture | Advanced digitalization - Enabling technologies |
| Call | Advanced and innovative digitalization 2022 |
Important results from the project
The goal of the project was to offer digital decision support tools to circular businesses in the furniture sector. These tools will help overcome the additional time and logistical cost of the ´backwards´ flow of resources in the circular economy using image recognition and machine learning. These prototype tools partially address the challenges of each stakeholder, however we encountered obstacles related to AI development as well as positive new ways to reframe e-commerce for circularity.
Expected long term effects
The project offers methodological approaches for AI tool development in the context of anomaly detection, working in dialogue with an large language model (LLM) to, first, understand what the model "sees", developing prompts that take advantage of this understanding to better identify damage, and then training the model to see specific types of damage. Practically, the project reveals evidence about the importance of customers´ pre-existing shopping practices when designing circular offerings.
Approach and implementation
The project encountered multiple unexpected setbacks beginning with one critical partner leaving the project after six months. This required a major reorientation of our plans and the recruitment of a new project partner (NORNORM). We also encountered multiple obstacles to the development of AI tools beginning with lack of existing sofa image data, changes in EU intellectual property law, and challenges developing an original and reliable dataset of sofa images.