Revolutionizing Bridge Operation and Maintenance with LLM-based Agents: An Overview of Applications and Insights

TL;DR

The paper addresses the lag in intelligence in bridge operation and maintenance and investigates how LLM-based agents can transform core O&M tasks. It surveys the evolution of LLM-based agents and proposes a methodology combining distributed knowledge, structured knowledge via knowledge graphs, and multi-round dialog data to build domain-specific agents for bridges. A concrete agent framework is presented, including perception, processing, and scheduling layers, with components like RAG, CoT, and LangChain to realize autonomous operation. The authors discuss development directions, expected benefits, and practical challenges, emphasizing industry-wide impact and the need for ethical and professional safeguards.

Abstract

In various industrial fields of human social development, people have been exploring methods aimed at freeing human labor. Constructing LLM-based agents is considered to be one of the most effective tools to achieve this goal. Agent, as a kind of human-like intelligent entity with the ability of perception, planning, decision-making, and action, has created great production value in many fields. However, the bridge O&M field shows a relatively low level of intelligence compared to other industries. Nevertheless, the bridge O&M field has developed numerous intelligent inspection devices, machine learning algorithms, and autonomous evaluation and decision-making methods, which provide a feasible basis for breakthroughs in artificial intelligence in this field. The aim of this study is to explore the impact of AI bodies based on large-scale language models on the field of bridge O&M and to analyze the potential challenges and opportunities it brings to the core tasks of bridge O&M. Through in-depth research and analysis, this paper expects to provide a more comprehensive perspective for understanding the application of intelligentsia in this field.

Figures (10)

• Figure 1: Modern bridge digital inspection and monitoring technology. It allows observation of the environment in which the bridge is located, the loads on the structure, the changes in the structure, and the load response of the structure.
• Figure 2: The evolutionary tree of modern LLMs traces the development of language models in recent years and highlights some of the most well-known models. Summarized by yang2024harnessing. Models on the same branch have closer relationships. Transformer-based models are shown in non-grey colors: decoder-only models in the blue branch, encoder-only models in the pink branch, and encoder-decoder models in the green branch. The vertical position of the models on the timeline represents their release dates. Open-source models are represented by solid squares, while closed-source models are represented by hollow ones. The stacked bar plot in the bottom right corner shows the number of models from various companies and institutions.
• Figure 3: Illustration of the growth trend in the field of LLM-based autonomous agents. wang2024surveysummarizes the development of intelligibles by time and number of articles. They assign different colors to represent various agent categories. For example, a game agent aims to simulate a game-player, while a tool agent mainly focuses on tool using. For each time period, they provide a curated list of studies with diverse agent categories.
• Figure 4: Distributed knowledge can be used to train word embedding models or as external information for retrieval.
• Figure 5: Knowledge graphs as the most typical structured knowledge.