The potential of data-science approaches for improving data collection, exchange, and analysis in Life Cycle Assessment
Main Presenter: Magdalena Rusch
Co-Authors: Josef-Peter Schöggl Rupert J. Baumgartner
Life Cycle Assessment (LCA) is dependent on accurate and high-quality data from the entire product lifecycle. The lack of such product and process data can make LCAs inaccurate. Moreover, data-driven decisions based on these results can lead to adverse effects for the environment and society. At the same time, the volume, velocity, and variety of data and the importance of new data sources are increasing for various sustainability areas, including LCA. Also, data-science approaches become more relevant to facilitate the collection, exchange, and analysis of (sustainability) data along a product value chain. However, it is not sufficiently clear how data-science methods could be linked to LCA and used for facilitating data-driven sustainability decision making in a circular economy. Therefore, this research aims to investigate the potential of data-science approaches, e.g., new modes of data collection, data sharing and data processing for LCA.
First, a literature review and a validation process conducted by the authors’ research group (which consists of eight members with an academic background in sustainable product and supply chain management, sustainability assessment, and circular economy) was done. This yielded a list of common LCA data and data-driven sustainability decision situation problems. Second, this list was then shared with LCA and data-science experts. In two (online) focus group workshops, the interlinkages and advantages and disadvantages of data-science approaches for LCA were discussed.
This resulted in a matrix that shows sustainability decision problems and LCA data problems and how to ease them with different data-science methods. The data-driven sustainability decision situation problems were categorized in seven areas: data-driven culture, data collection, data analytics, data integration, data sharing, data visualization and others. For example, when it comes to data sharing, privacy-preserving analytics (like homomorphic encryption) could increase stakeholders’ willingness to share data along a product value chain. Homomorphic Encryption allows calculations on encrypted data, but sensitive information (that could harm a firms’ intellectual property rights) is not revealed. The main goal of such approaches is to allow and foster collaboration by increasing security aspects while minimizing risks of data sharing. Thus, firms’ up- and downstream collaboration and data exchange can be facilitated by using data-science methods. The results build a basis for further research and are useful to aid the user in the LCA data management process – ultimately to foster data-driven decision-making on product sustainability.