Data science to boost the circular economy

By Dr. Beatriz Royo, ZLC Associate Professor.

Leer versión en español

It is a truth universally acknowledged that for both economic and environmental reasons, we must create ‘circular economies’ which as far as practicable reuse and recycle the materials, from rare earths and precious metals to the more mundane textiles and plastics, that our goods are made of. ZLC is a partner in the OPTIWASTE project, which aims to apply the latest advances in data sciences to improving the economic and environmental performance of the circular economy.

Rarely in these inverted supply chains can the processes be economically self-sustaining, and there is also a real risk that the environmental impact, especially of transport-related CO₂ emissions, may outweigh the benefits of recycling.

In Spain, as in other countries, ‘eco-taxes’ establish that at least 65% of products placed on the market are being levied on an increasing range of goods (electronics, textiles, batteries, plastics, etc) to underwrite the collection process, while regulations are being introduced to mandate the participation of industries in end of life cycle management. The levy varies depending on the source of materials involved (individuals or businesses).

Inputs – end of life goods - arise in many thousands of locations – there are around 40,000 consumer ‘drop-off’ locations or temporary accumulation centres for recyclables in Spain which are expected to increase as new sort of products have to comply with this eco-tax. (Waste from businesses is counted as ‘consumer’ if the company brings it to the collection point – there are other arrangements, and higher scales of charges, for collection on request from business premises). Products have to be sorted, broken down, and the different materials despatched to, often very specialised, reprocessing centres, and obviously while both washing machines and smartphones may offer similar materials for recycling in the form of plastics, electronic circuitry and so on, the ways these are handled and processed are very different. Above these accumulation centres is a network of waste management operations (SCRAPS) which control operations and are responsible for applying the eco-taxes.

For the OPTIWASTE study, ZLC and Predictland coordinated by PRONET-ISE and lead by ALIA and TECNARA are looking at electrical and electronic equipment waste generated through individuals, private companies and retail chains.  The project has three objectives.

• In environmental and sustainability terms, to reduce the amount of CO₂ generated per kilogram of waste managed.
• Economically, to reduce the cost per kilogram for transport and final disposal, and also to reduce the extra costs associated with the introduction of new products which may be incurred because they have been designed for easier disassembly and recycling, rather than just for ease of manufacturing.
• In the circularity agenda, to improve the recyclability of managed wastes and their reintroduction to production processes: that includes improving the quality of the recyclates obtained, which in turn is partly dependent on optimising the logistics and custody cycle.

ZLC’s contribution will be firstly around optimising the locations of the SCRAP control centres. Second, we will be optimising the number of ‘containers’ at temporary collection centres. Thirdly we will be attempting to use more robust predictions of demand calculated by PREDICTLAND – it is an interesting feature that some waste stream volumes are hard to predict: there may be seasonality, for example consumers may dispose of old goods after Christmas when they have been given new models or there may be a spike in the disposal of cooling fans after summer is over; or old smartphones or other consumer electronics may be discarded when a new generation comes on the market.

We are able to apply the mass analysis of data compiled by PRONET-ISE. These databases on waste collections and transactions extend in some cases over 17 years. This will enable the study to use the latest in Artificial Intelligence, Machine Learning and deep learning to create algorithms to model demand forecasts and to optimise the logistics networks, not just for the current situation, but as the scope of the circular economy, and the participation of consumers and businesses, evolves.

For more information, contact Dr. Beatriz Royo, ZLC Associate Professor.