{"id":{"repo_id":"mcmaster","oai_identifier":"oai:macsphere.mcmaster.ca:11375/33320"},"canonical_url":"https://search.ndltd.org/etd/mcmaster/oai:macsphere.mcmaster.ca:11375/33320","repository":{"repo_id":"mcmaster","name":"McMaster University","base_url":"https://macsphere.mcmaster.ca/server/oai/request"},"display":{"title":"Sensor Fusion and Artificial Intelligence Control of Additive Manufacturing Process","abstract":"Laser powder bed fusion (LPBF) enables the production of complex metal components, but its broader qualification is limited by defect formation, process variability, and the absence of sensing architectures that translate in-situ observations into controller-compatible process-parameter decisions. This thesis develops, integrates, and evaluates a synchronized multimodal sensing and feed-forward artificial-intelligence control framework for LPBF. The work first reviews machine-learning applications in LPBF monitoring and control to identify the gap between offline defect classification and actionable process control. It then develops a dual-sensor monitoring platform that combines structure-borne acoustic emission with coaxial two-colour pyrometry. An initial acoustic-emission pilot on a commercial LPBF platform established the practical observability limits of build-plate-mounted sensing under changing geometry and build height. These findings motivated the development of a synchronized dual-sensor architecture implemented on a modifiable LPBF platform, supported by geometry-aware experiments, Archimedes-density measurements, and micro-computed tomography validation. The resulting data stream is used to construct a physics-informed feed-forward control architecture for Inconel 625. Sensor-specific latent representations are fused with process-structure-property inputs and defect-state representations to generate candidate laser-power and scan-speed updates. Candidate updates are evaluated through a decoder-based simulation stage before being made available to the machine-control interface, thereby distinguishing the demonstrated system from autonomous closed-loop actuation. The thesis contributes an experimentally validated sensing pathway, a synchronized multimodal LPBF dataset linked to geometry and ground truth, and a controller-compatible predictive architecture for defect-aware process-parameter recommendation. Overall, the work advances LPBF monitoring beyond isolated sensor interpretation toward an integrated framework for sensing, defect estimation, and predictive control.","abstract_html":"Laser powder bed fusion (LPBF) enables the production of complex metal components, but its broader qualification is limited by defect formation, process variability, and the absence of sensing architectures that translate in-situ observations into controller-compatible process-parameter decisions. This thesis develops, integrates, and evaluates a synchronized multimodal sensing and feed-forward artificial-intelligence control framework for LPBF. The work first reviews machine-learning applications in LPBF monitoring and control to identify the gap between offline defect classification and actionable process control. It then develops a dual-sensor monitoring platform that combines structure-borne acoustic emission with coaxial two-colour pyrometry. An initial acoustic-emission pilot on a commercial LPBF platform established the practical observability limits of build-plate-mounted sensing under changing geometry and build height. These findings motivated the development of a synchronized dual-sensor architecture implemented on a modifiable LPBF platform, supported by geometry-aware experiments, Archimedes-density measurements, and micro-computed tomography validation. The resulting data stream is used to construct a physics-informed feed-forward control architecture for Inconel 625. Sensor-specific latent representations are fused with process-structure-property inputs and defect-state representations to generate candidate laser-power and scan-speed updates. Candidate updates are evaluated through a decoder-based simulation stage before being made available to the machine-control interface, thereby distinguishing the demonstrated system from autonomous closed-loop actuation. The thesis contributes an experimentally validated sensing pathway, a synchronized multimodal LPBF dataset linked to geometry and ground truth, and a controller-compatible predictive architecture for defect-aware process-parameter recommendation. Overall, the work advances LPBF monitoring beyond isolated sensor interpretation toward an integrated framework for sensing, defect estimation, and predictive control.","abstract_has_math":false,"creators":["Boer, Jan"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":"Mechanical Engineering","school":null,"contributors":[],"advisors":["Elbestawi, Mo"],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-18T19:40:20Z","subjects":[],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://doi.org/10.71548/182"],"render_values":[{"text":"https://doi.org/10.71548/182","href":"https://doi.org/10.71548/182","code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/11375/33320","outbound_label":"Handle","outbound_source":"dc:identifier.uri","record_url":"https://hdl.handle.net/11375/33320","record_label":"Repository record","record_kind":"handle","record_source":"dc:identifier.uri","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://macsphere.mcmaster.ca/server/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Amacsphere.mcmaster.ca%3A11375%2F33320","prefix":"dim"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Elbestawi, Mo"]},{"key":"dc:contributor.department","label":"Department","values":["Mechanical Engineering"]},{"key":"dc:creator","label":"Author","values":["Boer, Jan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-07-06T15:36:52Z"]},{"key":"dc:date.issued","label":"Date","values":["2026"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/11375/33320","https://doi.org/10.71548/182"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Laser powder bed fusion (LPBF) enables the production of complex metal components, but its broader qualification is limited by defect formation, process variability, and the absence of sensing architectures that translate in-situ observations into controller-compatible process-parameter decisions. This thesis develops, integrates, and evaluates a synchronized multimodal sensing and feed-forward artificial-intelligence control framework for LPBF. The work first reviews machine-learning applications in LPBF monitoring and control to identify the gap between offline defect classification and actionable process control. It then develops a dual-sensor monitoring platform that combines structure-borne acoustic emission with coaxial two-colour pyrometry. An initial acoustic-emission pilot on a commercial LPBF platform established the practical observability limits of build-plate-mounted sensing under changing geometry and build height. These findings motivated the development of a synchronized dual-sensor architecture implemented on a modifiable LPBF platform, supported by geometry-aware experiments, Archimedes-density measurements, and micro-computed tomography validation. The resulting data stream is used to construct a physics-informed feed-forward control architecture for Inconel 625. Sensor-specific latent representations are fused with process-structure-property inputs and defect-state representations to generate candidate laser-power and scan-speed updates. Candidate updates are evaluated through a decoder-based simulation stage before being made available to the machine-control interface, thereby distinguishing the demonstrated system from autonomous closed-loop actuation. The thesis contributes an experimentally validated sensing pathway, a synchronized multimodal LPBF dataset linked to geometry and ground truth, and a controller-compatible predictive architecture for defect-aware process-parameter recommendation. Overall, the work advances LPBF monitoring beyond isolated sensor interpretation toward an integrated framework for sensing, defect estimation, and predictive control."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["Doctor of Philosophy (PhD)"]},{"key":"dc:title","label":"Title","values":["Sensor Fusion and Artificial Intelligence Control of Additive Manufacturing Process"]},{"key":"others:access_status","label":"Others Access Status","values":["open.access"]}]}],"canonical_facts":{"dc:contributor.advisor":["Elbestawi, Mo"],"dc:contributor.department":["Mechanical Engineering"],"dc:creator":["Boer, Jan"],"dc:date.accessioned":["2026-07-06T15:36:52Z"],"dc:date.issued":["2026"],"dc:description.abstract":["Laser powder bed fusion (LPBF) enables the production of complex metal components, but its broader qualification is limited by defect formation, process variability, and the absence of sensing architectures that translate in-situ observations into controller-compatible process-parameter decisions. This thesis develops, integrates, and evaluates a synchronized multimodal sensing and feed-forward artificial-intelligence control framework for LPBF. The work first reviews machine-learning applications in LPBF monitoring and control to identify the gap between offline defect classification and actionable process control. It then develops a dual-sensor monitoring platform that combines structure-borne acoustic emission with coaxial two-colour pyrometry. An initial acoustic-emission pilot on a commercial LPBF platform established the practical observability limits of build-plate-mounted sensing under changing geometry and build height. These findings motivated the development of a synchronized dual-sensor architecture implemented on a modifiable LPBF platform, supported by geometry-aware experiments, Archimedes-density measurements, and micro-computed tomography validation. The resulting data stream is used to construct a physics-informed feed-forward control architecture for Inconel 625. Sensor-specific latent representations are fused with process-structure-property inputs and defect-state representations to generate candidate laser-power and scan-speed updates. Candidate updates are evaluated through a decoder-based simulation stage before being made available to the machine-control interface, thereby distinguishing the demonstrated system from autonomous closed-loop actuation. The thesis contributes an experimentally validated sensing pathway, a synchronized multimodal LPBF dataset linked to geometry and ground truth, and a controller-compatible predictive architecture for defect-aware process-parameter recommendation. Overall, the work advances LPBF monitoring beyond isolated sensor interpretation toward an integrated framework for sensing, defect estimation, and predictive control."],"dc:description.degree":["Doctor of Philosophy (PhD)"],"dc:identifier.uri":["https://hdl.handle.net/11375/33320","https://doi.org/10.71548/182"],"dc:title":["Sensor Fusion and Artificial Intelligence Control of Additive Manufacturing Process"],"dc:type":["Thesis"],"others:access_status":["open.access"]},"updated_at":"2026-11-18T19:40:20Z"}