• Home
  • NEWS
  • News
  • REMAIN project moves forward with the first scientific publications and an IA-based defect detection dataset

REMAIN project moves forward with the first scientific publications and an IA-based defect detection dataset

The partners of the European project coordinated by Inescop gathered at a transnational meeting held at the University of Coimbra, Portugal, to share the progress made in research and plan the actions for 2025 in the framework of a project that seeks to promote remanufacturing in industrial manufacturing sectors, such as footwear.

The REMAIN project is progressing with the first scientific publications on its technological advances and the proposal to create a database of images for breaks and defects detected in footwear, which, suitably labelled, will form the basis to train AI. The idea of compiling this dataset was finalised during the transnational meeting held in Coimbra on 31st January. The gathering was attended by the partners of the European project REMAIN (Robotic REMAnufacturing of Deformable INdustrial Products), which is led by Inescop as the Spanish Footwear Technology Centre.

ReunionRemain COIMBRA1 07 02 25

This initiative also involves the Federation of Spanish Footwear Industries (FICE), AIDA CCI, the Universities of Alicante, Clermont Auvergne INP, Coimbra and Zaragoza; as well as the Spanish company Automática y Control Numérico SL, the French company SMA-RTY, and the social association Proyecto Lázaro as a collaborator.

The meeting took place at the Institute of Systems and Robots at the University of Coimbra, where discussions centred on the systems and methodology to be applied in the REMAIN project to detect damage in consumer products with the aim of promoting their remanufacturing. In this sense, the experts analysed the possible conventional detection systems and other AI-based methods that can be applied from labelled databases. Meanwhile, the role of the meta-sensor that is being developed in the project was presented, which aims to locate and identify the product and its damage, using various machine vision technologies: RGB-D, NIR, SWIR, TOF. This use is intended to be linked both to the initial identification of the defect and to robotic perception and control, in order to recognise objects, track them and serve as a basis for their manipulation, through mono-robotic or bi-robotic systems.

ReunionRemain COIMBRA2 07 02 25

The experts gathered at the meeting also identified the most common types of damage that occur to consumer goods, mainly footwear and textiles, and analysed the production processes that are most feasible for the possible repair or remanufacturing of these goods. The ultimate goal is to develop an eco-design guide that offers guidelines appropriate to the new remanufacturing paradigm. The last part of the transnational meeting focused on communication and awareness-raising tasks, including the results of a market study focused on consumers and companies, the production of an informative video and the presentation of scientific advances in publications and events.

ReunionRemain COIMBRA3 07 02 25

In short, the aim of this European project is to promote the remanufacture and reintroduction of products onto the market in order to extend their life cycle, especially in sectors with a strong presence in the SUDOE region: the footwear, textile and furniture industries. This European action is part of the Interreg Sudoe programme, the territorial cooperation programme for south-west Europe, which supports regional development by co-funding transnational projects through the European Regional Development Fund (ERDF).

 
Share this