A multi-machine application connecting field measurements to energy monitoring, model-based prediction, and future energy-saving decisions.
Highlights
- Integrated CNC states, CT measurements, smart meters, and model services into a multi-machine monitoring application.
- Implemented state-based power-factor substitution to support energy estimation after meter removal.
- Connected field data to energy-prediction APIs and established configurable control-point mapping for future energy-saving functions.
Purpose
- Establish a path from equipment measurement and energy prediction to energy-saving decisions.
- Collect reference meter data during model development, then support ongoing monitoring after the meter is removed.
Implementation
- Analyzed machine subsystems, measurement locations, phase assignments, channel scaling, voltage, and controller states.
- Integrated CT sensors, Remote I/O, smart meters, and CNC/PLC interfaces into the monitoring application.
- Gave each machine independent settings, connections, measurement data, trends, logs, and service lifecycle management.
- Collected power-factor statistics by operating state and supported their use after reference-meter removal.
- Prepared model inputs and integrated team-provided HTTP prediction services, displaying and recording energy-prediction results.
- Established configurable control-point mapping, test reads and writes, throttling, and pre-write checks for future control integration.
Delivery and Next Step
- Deployed core monitoring at two sites, with approximately one year of operation during my involvement.
- Completed equipment integration, the monitoring application, data processing, power-factor substitution, and model API integration; model training and prediction servers were team-provided.
- Current model output covers energy prediction. The planned extension adds recommendations for switching off auxiliary modules, with equipment conditions and safety validation required before automatic control.