bridge-inspection

该论文提出一种车辆集成的数字孪生框架，通过驱动感知（drive-by sensing）实现桥梁结构健康监测，结合傅里叶神经算子代理模型、贝叶斯优化的传感器设计以及无监督损伤评估，并在澳大利亚和日本的多站点现场试验中完成验证。

This paper proposes a Damage Cause Encoder for bridge diagnostic agents that uses knowledge triple extraction from manuals and retrieval-augmented fine-tuning with QLoRA, achieving high accuracy with lower memory usage.

An essay expanding on the metaphor of open source as critical infrastructure by examining how real-world bridges are inspected and maintained, drawing parallels to the need for systematic, government-backed support for open source projects.