Encoding Invisible Causation for Bridge Diagnostic Agents: Triple-Guided Retrieval-Augmented Fine-Tuning with QLoRA

A new Damage Cause Encoder proposed by researchers achieves 87.07% test accuracy in classifying 10 damage causes from visible bridge descriptions by chaining knowledge triple extraction, retrieval-augmented context, and QLoRA fine-tuning. QLoRA matches full-precision LoRA accuracy while delivering 11% faster inference and 72% lower GPU memory, enabling deployment on consumer-grade hardware.

arXiv:2607.21680v1 Announce Type: new Abstract: Bridge infrastructure deteriorates gradually, yet its root causes---salt intrusion, freezing, fatigue cracking, and others---remain invisible to the naked eye. Expert diagnosis relies on tacit knowledge built over years of practice. We address the challenge of automating this latent causal reasoning by proposing a Damage Cause Encoder that classifies 10-class damage causes from visible damage descriptions $S i$ for use in autonomous bridge diagnostic agents. Our approach chains three components: i Knowledge Triple Extraction---a large language model extracts causal triples of the form damage $\xrightarrow{\mathtt{caused\ by}}$ cause from 15--35 diagnostic PDF manuals and indexes them in a FAISS vector store; ii Retrieval-Augmented Context---at training and inference time, relevant causal triples $\mathcal{C} i$ are retrieved and concatenated with $S i$, converting implicit domain knowledge into explicit Encoder context; iii Systematic Fine-tuning Comparison---we conduct a rigorous comparison of LoRA, QLoRA, and QA-LoRA on a fixed Golden Testset 116 stratified samples , demonstrating that QLoRA achieves the optimal trade-off: identical test accuracy 87.07% to full-precision LoRA, 11% faster inference, 72% lower GPU memory, and superior generalization across diverse unseen inputs. A controlled Golden Testset---stratified, deduplicated, and difficulty-tagged---is introduced as a reusable benchmark contribution. QLoRA further outperforms LoRA by 13 percentage points on a 100-sample diverse evaluation spanning all 10 damage cause classes.These findings enable memory-efficient, high-accuracy diagnostic agents on consumer-grade hardware for edge deployment.