| Proj No. | A3254-251 |
| Title | LLM-Powered Industrial Anomaly Detection for Predictive Maintenance |
| Summary | Objective This project aims to develop an LLM-enhanced industrial anomaly detection system that analyzes sensor data from industrial machines to predict potential failures and suggest preventive maintenance strategies. Scope Use public industrial sensor datasets (e.g., Hydraulic System Condition Monitoring Dataset) containing temperature, pressure, vibration, and power usage data. Apply IoT-LLM techniques to transform raw sensor data into LLM-friendly descriptions for anomaly detection. Utilize LLM reasoning with Retrieval-Augmented Generation (RAG) to generate explainable maintenance reports predicting failures and suggesting actions. Required Skills Python, Pandas, and basic machine learning (Scikit-learn, PyTorch). Understanding of sensor data processing and industrial maintenance concepts. Experience with OpenAI API / LLaMA / Mistral for LLM-enhanced analytics. |
| Supervisor | Ast/P Yang Jianfei (Loc:School of Mechanical and Aerospace Engineering (MAE), Ext: ?) |
| Co-Supervisor | - |
| RI Co-Supervisor | - |
| Lab | Internet of Things Laboratory (Loc: S1-B4c-14, ext: 5470/5475) |
| Single/Group: | Single |
| Area: | Digital Media Processing and Computer Engineering |
| ISP/RI/SMP/SCP?: |