Analytics-Driven Cooperative Multi-Robot Additive Manufacturing: Advancing Efficiency and Quality

Robotic additive manufacturing (AM) leverages the unbounded work envelopes and high degrees of freedom of articulated manipulators to revolutionize large-scale fabrication; however, the transition from single robot to cooperative multi-robot and multi-head systems introduces complex, coupled decisions in decomposition, placement, and path planning under time, energy, quality, strength, and thermal constraints. This dissertation establishes a comprehensive analytics framework that links robot kinematics and dynamics, process parameters, and cooperation strategies to the resulting process–structure–property relationships in multi-robot AM. First, it characterizes the dimensional accuracy and energy consumption of single-robot AM through analytically grounded and experimentally validated model and use these models to construct three-dimensional energy–quality (EQ) maps that relate workspace location to manufacturing performance and guide part placement within a manipulator’s workspace. Second, to address multi-robot scenarios, these concepts of EQ maps are extended into inverse EQ maps for multi- robot scenarios, enabling the optimal positioning of multiple robot bases around decomposed sub-parts while satisfying reachability, collision, and tolerance constraints to enable cooperative printing. Third, to accelerate decision-making and address the computational limitations of traditional analytical models, the thesis introduces a Kinematics-Guided Multi-Task learning architecture that jointly predicts path feasibility and energy consumption. By embedding inverse kinematics into a shared backbone, KG-MT internalizes reachability and joint limits, achieves orders-of-magnitude speedups over simulation, and generalizes across robot platforms to support cross-platform, energy-aware planning. Finally, the thesis addresses the critical process planning challenges of strength- and thermal-aware decomposition and path planning respectively, : for multi-head extrusion-based systems, where decomposition creates weak interfaces, thesis proposes a Synchronous Multi-Layer Printing strategy that recovers the joint strength of single-head deposition while preserving the throughput of multi-head printing; for multi-laser systems, the thesis establishes a framework that links decomposed multi-laser trajectories to thermal histories and uses clustering-based selection of path-planning strategies to stabilize thermal distributions and promote uniform temperature fields, validated experimentally and via a physics-informed neural network. Collectively, these contributions unify robot-, process-, structure-, and property-level analytics into a coherent decision-making foundation for cooperative multi-robot AM, supporting energy-efficient, high- quality, and property-aware decision-making across placement, decomposition, and path planning.

Author dc:creator

• Suyog Ghungrad (24400076)

• Engineering, Industrial
• Engineering, Mechanical
• Computer Science

Statement dc:rights

• In Copyright
• Open Access after 2028-05-01

DOI dc:identifier

https://doi.org/10.25417/uic.32995121.v1

OAI identifier oai:identifier

oai:figshare.com:article/32995121