{"paper":{"title":"Few-Shot Physics-Informed Neural Network for Shape Reconstruction of Concentric-Tube Robots","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"Embedding Cosserat rod equations in a neural network enables accurate full-state reconstruction of concentric-tube robots from few-shot data.","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Filipe C. Pedrosa, Jagadeesan Jayender, Navid Feizi, Rajni V. Patel","submitted_at":"2026-05-12T22:07:53Z","abstract_excerpt":"Modeling concentric tube robots (CTRs) involves complex nonlinear continuum mechanics, and despite recent progress, physics-based models often lack an accurate representation of the experimental setups. To overcome these limitations, deep neural network-based models have been explored as alternatives with superior accuracy; however, they often overlook known mechanics, require large training datasets, and typically discard shape estimation of the robot. We present a physics-informed neural network (PINN) for kinematic modeling of a 6-DoF CTR with three pre-curved tubes that embeds the Cosserat"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"PINN enables full-state estimation of shape, twist angle, torsional strain, bending moment, and orientation. Benchmark tests show a mean shape error below 1% of the robot length and accurately recovered other kinematic states, outperforming a purely physics-based Cosserat rod model baseline while using a minimal training set.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That embedding the Cosserat rod differential equations directly into the neural network loss will produce accurate full-state estimates from few-shot data without requiring large datasets or suffering from physics-data mismatch in real experimental setups.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A PINN embedding Cosserat rod mechanics achieves sub-1% mean shape error for 6-DoF concentric tube robots using minimal training data and outperforms a pure physics baseline.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Embedding Cosserat rod equations in a neural network enables accurate full-state reconstruction of concentric-tube robots from few-shot data.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"35ca0b1a95583f7ad9b6adc345502e13ac7033b8887922b8e8d4097172d8b145"},"source":{"id":"2605.12790","kind":"arxiv","version":1},"verdict":{"id":"60ad6628-f612-4c29-bcb5-80fb3d541fbd","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:37:13.379849Z","strongest_claim":"PINN enables full-state estimation of shape, twist angle, torsional strain, bending moment, and orientation. Benchmark tests show a mean shape error below 1% of the robot length and accurately recovered other kinematic states, outperforming a purely physics-based Cosserat rod model baseline while using a minimal training set.","one_line_summary":"A PINN embedding Cosserat rod mechanics achieves sub-1% mean shape error for 6-DoF concentric tube robots using minimal training data and outperforms a pure physics baseline.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That embedding the Cosserat rod differential equations directly into the neural network loss will produce accurate full-state estimates from few-shot data without requiring large datasets or suffering from physics-data mismatch in real experimental setups.","pith_extraction_headline":"Embedding Cosserat rod equations in a neural network enables accurate full-state reconstruction of concentric-tube robots from few-shot data."},"references":{"count":36,"sample":[{"doi":"","year":2022,"title":"Continuum robots for medical interventions,","work_id":"e8f2d8f1-0c0a-4291-a066-2691812966a8","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2015,"title":"Continuum robots for medical applications: A survey,","work_id":"3faa63ec-ae1e-458d-932a-62ca2aef8f40","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Design and fabrication of concentric tube robots: A survey,","work_id":"bd297905-8a22-4391-b0db-c953b9a03b61","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"From theoretical work to clinical translation: Progress in concentric tube robots,","work_id":"65066cff-eed9-4aa6-bb48-5b3442dccd37","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Using robotics to move a neurosurgeon’s hands to the tip of their endoscope,","work_id":"108e4e5f-09ce-4639-9ab8-bf5e21c7daa0","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":36,"snapshot_sha256":"14c2cf0210d7b4e8e318dcbd052e5a5ddd301689c37266edd472d0725f42a34d","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}