Company or Institution
Sustainable Development Goals (SDGs)
SDG 9: Industry, Innovation and Infrastructure
SDG 12: Responsible Consumption and Production
General description of the AI solution
Consumer goods account for almost 60% of global environmental impact and involve hundreds of millions of workers around the world (Ivanova & al, 2016). Manufactured, used by the consumer and then disposed of, they generate numerous socio-environmental externalities throughout their life cycle, prompting the United Nations to make “Responsible Consumption and Production” one of the SDGs in 2012.
To tackle this goal, reliable and product-specific information are necessary for both producers, strengthening their decision-making, and for consumers, enabling them to account for sustainability in their purchases. In order to obtain such data rigorously, the ISO-standardized life cycle assessment (LCA) has been developed for the last two decades, providing a framework to quantify impacts from cradle-to-grave. However, while it is currently the most robust methodology, conducting a multi-criteria study of each process involved from the extraction of raw materials to elimination, it also proves to be tedious and costly. As such, despite an increasing demand from companies and LCA’s inclusion in several French and European regulations, there are no solutions to evaluate impacts without compromising on details.
By leveraging AI on key aspects of LCA – data collection, modelling, and recommendations – we achieve unprecedented levels of efficiency and empower brands to assess their entire product catalogs (1000+ SKU), accelerating the transition from a paradigm of punctual assessment to one of continuous improvement. Thanks to dedicated NLP algorithms, we enhance traceability by automating stakeholders identification (Knowledge Graph) and data extraction (invoices, scope and transaction certificates for labels). To grant novices the expertise of LCA analysts, we are training a Graph ML model to both detect inconsistencies in the modelling and predict parts of the supply chain. Finally, we implemented supply chain optimisation to generate alternative scenarios, accounting for sustainability and economic costs. Blended within our platform, these AI tools foster more responsible products.
Github, open data repository, prototype or working demo
Winner of the Big Data & AI for Good – September 2022
Winner of the AI for Tomorrow Challenge – April 2022