The circular economy has become an increasingly fashionable concept intended to help with the environmental and economic issues that the traditional linear supply chain created. Against this backdrop, circular economic integration in supply chain management practice has become crucial for sustainability. Part of this integration is introducing predictive maintenance strategies in line with the circular economy goal of resource efficiency and waste minimization. Predictive maintenance uses data-driven models to predict when equipment or components are going to fail thereby facilitating remediation that may extend life assets and waste minimization. This approach contributes to less premature replacement of parts and equipment in a circular economic context as way restorative supply chains protecting valuable resources and minimizing impact of production. This paper proposes an innovative framework targeted at circular predictive maintenance processes using multi-agent-based presentation to expand the life cycle of the machines and enhance circular and sustainable supply chain management.