Towards Multi-Turn Dialog Systems for Industrial Asset Operations and Maintenance
The paper presents a multi-turn dialog system tailored for industrial asset operations and maintenance. This system addresses limitations of traditional architectures by enhancing context maintenance and tool invocation efficiency. Evaluation results indicate significant improvements in response quality and task completion rates compared to baseline methods.
- ▪The proposed system utilizes a supervisor-specialist multi-agent architecture.
- ▪It incorporates structured artifact reuse, dynamic replanning, and parallel tool execution to improve efficiency.
- ▪Evaluation shows a 54.5% increase in planning effectiveness and a 37.8% improvement in task completion.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.24953 (cs) [Submitted on 24 May 2026] Title:Towards Multi-Turn Dialog Systems for Industrial Asset Operations and Maintenance Authors:Chengrui Li, Rujing Li, Yitong Bai, Rui Li View a PDF of the paper titled Towards Multi-Turn Dialog Systems for Industrial Asset Operations and Maintenance, by Chengrui Li and 3 other authors View PDF HTML (experimental) Abstract:Industrial asset operations and maintenance question answering is inherently multi-turn, iterative, and highly dependent on external tool invocation. However, the conventional plan-execute single-agent architecture exhibits clear limitations in maintaining cross-turn context, and reusing intermediate results.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
