Development of a Scalable Predictive Maintenance Model for Industrial Equipment using Time-Series Data

We are seeking an experienced machine learning engineer to develop a predictive maintenance model for our fleet of industrial robots and CNC machines. The goal is to minimize downtime, optimize maintenance schedules, and reduce overall equipment costs. This project will involve data collection from various sensors (vibration, temperature, pressure), data preprocessing and cleaning using Python libraries like Pandas and NumPy, feature engineering to extract relevant information from the time-series data, training a machine learning model (ideally using TensorFlow or PyTorch) to predict equipment failures based on historical data and real-time sensor readings, evaluating model performance using appropriate metrics (precision, recall, F1-score), and deploying the model to a cloud platform for real-time predictions. We require strong experience in time series analysis, anomaly detection, and model deployment. A key component will be building a robust data pipeline that can handle high volumes of sensor data reliably. We are open to exploring different model architectures (e.g., LSTM, GRU, Transformers) and evaluation methods. Post-deployment support for model monitoring and refinement will be required for a period of 3 months. Experience with cloud platforms like AWS or Azure is preferred. Documentation and API access will be provided.