Data for circularity assessment: what, where, and how?

The growing need for circular economy (CE) assessment in manufacturing enterprises is driven by increasing regulatory and market pressures. However, existing approaches provide limited support for identifying and collecting the required data to compute circularity indicators. The main contribution of this paper is a structured method that links circularity indicators to collectable data and formalizes data collection as a decision problem. CE assessment is operationalized through a four-phase, data-driven approach comprising (1) an industrial diagnosis to define objectives, scope, and relevant information systems, (2) a systematic sense-making process to translate CE assessment resources into structured data requirements, (3) a decomposition of circularity indicators into data points, referred to as c-data, and (4) a mixed-integer linear programming model to determine an efficient data collection strategy. Applied to a medical radiology system manufacturer case study, 69 target indicators are decomposed into 91 c-data points distributed across eight information systems. Building on this, the optimization model shows that accessing only a limited subset of systems can be sufficient to compute a substantial share of indicators, thereby revealing trade-offs between implementation effort and assessment coverage. In this context, three alternative data collection strategies are identified for the company. Notably, one of these strategies demonstrates that accessing just one of eight information systems makes it possible to collect 13% of eligible data points, which in turn enables the computation of 25% of CE indicators. The approach supports the transition from high-level assessment requirements to actionable data sourcing strategies and is extendable to other sustainability assessment contexts.