Disassembly Learning Catalyst (DLC)

Technology Summary

This technology employs enhanced multilayer direct disassembly networks (MDDNs) derived from CAD assemblies to model and optimize product disassembly sequences. Using knowledge graphs converted from MDDNs, a generative adversarial network (GAN) framework with generative and discriminator components is trained to produce synthetic disassembly designs that improve component accessibility for efficient disassembly, especially considering component degradation. The system integrates collision detection, interference analysis, and node-edge attribute augmentation to create detailed disassembly network representations, enabling cost-effective and feasible disassembly planning and automated CAD design refinement focused on target parts

Key Advantages

• Generates optimized disassembly sequences considering complex part interferences and clustering within CAD assemblies.
• Embeds physics-based disassembly simulations into AI architecture for simulating and generating disassembly plans and disassemblability insights from CAD design features.
• Incorporates component degradation effects into disassembly network simulations for realistic end-of-life planning.
• Uses generative adversarial learning to iteratively improve product designs that ease disassembly and reduce costs.
• Enables hierarchical multilayer network modeling bridging assembly level and sub-assembly levels for detailed analysis.
• Integrates user feedback for iterative design refinement, ensuring practical usability and customization.
• Transforms abstract disassembly networks into manufacturable CAD assembly files for direct application.
• Reduces difficulty in planning cost-effective and feasible disassembly sequences for complex assemblies.
• Reduces challenges in designing disassemblability accounting for component degradation and wear.
• Limits manual and time-consuming disassembly planning and CAD redesign processes lacking automation.
• Addresses gap between abstract disassembly planning models and practical CAD assembly designs.

Market Opportunities

• Automotive and aerospace assembly and component designs optimized for reuse, maintenance, remanufacturing, and recycling.
• Electronics and consumer product manufacturing with easier repair and end-of-life disassembly.
• Industrial machinery and equipment design for maintenance and modular upgrades.
• CAD and Simulation software platforms incorporating AI-driven disassembly optimization tools.
• Product lifecycle management systems focused on sustainability and cost reduction.
• Robotics and automation systems for disassembly and recycling line operations.

Stage of Development

Proof of Concept

Patent Status

Pending

  References & Publications

2022 Joao Paulo Jacomini Prioli, Header M. Alrufaif , Jeremy L. Rickli Disassembly assessment from CAD-based collision evaluation for sequence planning Robotics and Computer-Integrated Manufacturing December 2022, 102416

2023 Alrufaifi, H. M., Prioli, J. P. J., & Rickli, J. L. (2023, June). Degradation-Based Design for Disassembly Assessment Using Network Centrality Metrics. In International Conference on Flexible Automation and Intelligent Manufacturing (pp. 736-744). Cham: Springer Nature Switzerland.

2025 Varupala, S. S. V. P., Prioli, J. P., & Rickli, J. L. (2025, August). Graph Neural Networks for Interference Matrix Prediction in Generative Design for Disassembly. In International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (Vol. 89244, p. V004T05A016). American Society of Mechanical Engineers.