Product Digital Twin Supporting End-of-life Phase of Electric Vehicle Batteries Utilizing Product-Process-Resource Asset Network

TL;DR

The paper tackles end-of-life management for electric vehicle batteries within a circular economy by proposing a Product Digital Twin (PDT) framework built on a Bi-Flow PAN (Bi-PAN) that unifies product data, manufacturing processes, and resources for both assembly and disassembly. It introduces Bi-PAN as an extension of the PAN model to support bidirectional life-cycle flows, enabling planning, optimization, and data-driven decision-making across remanufacturing and recycling workflows. The approach is demonstrated on an EV battery use-case, leveraging AutomationML and aligning with industry standards to capture component dependencies and process resources. The study shows that Bi-PAN enhances end-of-life planning, supports remanufacturing and recycling, and offers a pathway toward more sustainable and circular battery life-cycle management, with future work focused on deeper integration with Asset Administration Shell and scalability to full battery PAN models.

Abstract

In a circular economy, products in their end-of-life phase should be either remanufactured or recycled. Both of these processes are crucial for sustainability and environmental conservation. However, manufacturers frequently do not support these processes enough in terms of not sharing relevant data about the products nor their (re-)manufacturing processes. This paper proposes to accompany each product with a digital twin technology, specifically the Product Digital Twin (PDT), which can carry information for facilitating and optimizing production and remanufacturing processes. This paper introduces a knowledge representation called Bi-Flow Product-Process-Resource Asset Network (Bi-PAN). Bi-PAN extends a well-proven Product-Process-Resource Asset Network (PAN) paradigm by integrating both assembly and disassembly workflows into a single information model. Such networks enable capturing relevant relationships across products, production resources, manufacturing processes, and specific production operations that have to be done in the manufacturing phase of a product. The proposed approach is demonstrated in a use-case of disassembling electric vehicle (EV) batteries. By utilizing PDTs with Bi-PAN knowledge models, challenges associated with disassembling of EV batteries can be solved flexibly and efficiently for various battery types, enhancing the sustainability of the EV battery life-cycle management.

Figures (4)

• Figure 1: Bi-PAN for the simplified BMW i3 EV battery, representing product stages, production process operations, and resource skills. The legend explaining the nomenclature is included in the bottom part of the figure.
• Figure 2: Example of a system unit class representing the BMW i3 battery in AutomationML as a screenshot from the AutomationML Editor.
• Figure 3: Changing a broken module use-case. In the generic Bi-PAN (depicted in Fig. \ref{['figPAN']}), the module to be switched is found, the proper forward or backward operations are selected and the whole production recipe is extracted for this specific reparation case. Whereas the generic Bi-PAN has bidirectional arrows, this specific procedure extracted for the selected case has already single-directional arrows that can be already executed by the manufacturing system.
• Figure 4: Evaluation of the digital twin design approaches based on Likert's scale: (i) the conventional DT design on the left and (ii) Bi-PAN-based DT design on the right.