Vision feedback on freeform-shaped structure for 3D printing

• Main Author: Quah, Noel Tan Kai
• Other Authors: Li King Ho Holden
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Computer vision; Feedback control; Image segmentation; Robotics; 3D printing; Piezoelectric sensors; Electromechanical impedance; In-process restoration
• Online Access: https://hdl.handle.net/10356/180473
• Institution: Nanyang Technological University

In Singapore, there is an increasing need for independence in manpower within the Building and Construction (B&C) Industry. The current construction approach in Singapore’s high density urban landscape is the use of Prefabricated Prefinished Volumetric Construction (PPVC), where 3-dimensional modules with completed finishes, fixtures, and fittings are fabricated in a decentralised facility prior to its delivery and designated on-site installation. This construction method has benefited the populous with reduced environmental pollutants and reduced reliance to on-site personnel, while lowering exposure to construction site health hazards. Moreover, smaller module production permits improved productivity, quality control, and customizability. The highly modular requirements for PPVC places additive manufacturing in an advantageous position due to its high customisability, and low volume manufacturing capabilities for faster manufacturing response time, faster production changeovers, and lower inventory requirements. Freeform structures, characterized by their complex, non-linear, and organic shapes, presents itself as an opportunity in 3D Concrete Printing (3DCP) by opening new emerging technological advances such as customizable architectural design and material science advancements. As demand for more intricate and personalized design increases, the need to be autonomous, precise, and fail-proof becomes increasingly critical. Hence, the need for monitoring systems for 3D printing technologies becomes a growing priority as 3DCP technologies move towards autonomous maturity. The discipline of vision feedback for 3DCP is a field that stands at the intersection of imaging technologies, material engineering, and process control. While the inherent characteristics of autonomous 3D printing may seem advantageous, 3DCP has only begun to move away from early-stage development and the success rate in the real-world environment is still being evaluated as buildability, extrudability, and pumpability concerns persists. The objective of this thesis is to develop a series of methods that lead to the development of a visual feedback system for freeform-shaped objects in 3DCP. A series of studies were conducted to assess the feasibility of the experiment prior to the stated aim. This includes the assessment and development of imaging technologies using computer vision tools that are assessed with case studies, a feedback mechanism with sensor implementation, and a real-time process control solution for 3DCP.