(345d) i4metal: Innovative Data Science Technologies for Metal Recovery from Scrap

The recycling of metals is significantly more cost- and resource- efficient compared to throwing the metals in a landfill and then recover them. There has long been an effort to recover metals and subsequently reuse them in new useful products. As a consequence, metal recycling technologies are mature enough¹ and are expected to play a fundamental role in the transition to circular economy² over the next decades. Scrap refers to materials that are by-products or waste from industrial activity but also refers to end-of-life products, such as end-of-life vehicles (ELVs), waste electric and electronic equipment (WEEE), used beverage cans (UBCs), etc., whose use spans all aspects of common human activities.

The importance of metal recycling³ and the large number of materials, among other reasons, explain why there is a large number of companies whose activities are within the recycling cycle of these materials. Trading and treatment companies receive scrap from collector companies, recover the metal by physical processes such as dismantling, shredding, etc., and then sell it to metal refining industries. Trading and treatment companies face significant challenges such as:

• Difficulty in identification of materials and classification according to international standards. This is usually done manually with low productivity. Central role in the effectiveness of this process plays the tacit knowledge of the involved employees, to whom the whole process heavily relies on.
• The confrontation with malign activities by scrap supplier. It is commonly observed that suppliers, in an effort to maximize their profit, mix several polluting materials with the useful scrap.
• Minimization of the environmental footprint. As every industry, metal scrap trading and treatment companies need to operate in the most effective way that will minimize the environmental cost which comes as a consequence of their activities.

i4metal, is a research program co-financed by the Operational Program Competitiveness, Entrepreneurship and Innovation, under the call RESEARCH-CREATE-INNOVATE, is implemented in a partnership among ATHENA research center, National Technical University of Athens and ANAMET S.A.. The goal of the project is the development of information and communication technologies (ICT), that will facilitate several aspects of common processes in ANAMET S.A.. These include:

• The improvement of the quality control of the scrap that ANAMET S.A. manages.
• The development of knowledge graphs for the representation of the qualitative and quantitative characteristics of the materials and several processes involved in their treatment.
• The easier identification of malign actions.
• The optimization of the operation of the automobile shredder of ANAMET S.A..

In this poster the development of two software applications as part of the project will be presented:

(a) a knowledge graph⁴ that includes qualitative and quantitative information about scrap materials. Alternative processes associated with the treatment and final refinement of scrap metals will be modelled. This knowledge currently exists in an unstructured form, thus it poses a challenge for its exploitation in decision making. A methodology for incorporation of the tacit knowledge of ANAMET S.A. employees will also be discussed. The graph will be used for the identification of strong correlation among its data by applying relevant queries.

(b) an application aiming to classify scrap metal shipments in terms of their precarity with regard to malign actions. This application will be based on machine learning algorithms and is anticipated to help the company to minimize time and cost to shipments that are less likely to comprise foreign materials.

References

¹M. Reuter et al., “Metal recycling. Opportunities, Limits, Infastruture”, United Nations Environment Programme, Editor: International Resource Panel, Working Group on the Global Metal Flows, 2013

²D.A.R. George, B. C-A Lin, Y. Chen, “A Circular economy model of economic growth”, Environmental, Modelling & Software, Vol. 73, 2015, p. 60-63

³C. Hageluken et al., “The EU circular economy and its relevance to metal recycling”, Recycling, 2016, 1, 242-253

⁴N. Noy et al., “Industry-scale knowledge graphs: Lessons and challenges”, acm queue, Vol. 17, Issue 2 , 2019