acm-master

You can orchestrate an automated condition monitoring pipeline by running batch jobs, managing coldstarts, and deploying detectors to monitor industrial equipment like FD_FAN and GAS_TURBINE assets. This unifies data ingestion, multi-detector analytics, and RUL forecasting into a scalable architecture.

You can forecast remaining useful life by running RUL forecasting tasks within the ACM pipeline, utilizing a configured SQL Server backend for data management. The pipeline handles the end-to-end model lifecycle, from fusion configuration to model promotion.

Yes, you can visualize condition monitoring results in Grafana. The framework provides observability through Grafana dashboards and integrates traces and logs from the observability stack, including Tempo, Loki, and Prometheus, to monitor pipeline runs.

Yes, you need a Python-based ACM codebase, a configured SQL Server backend, and access to the observability stack to run the workflows. Batch runner tooling and T-SQL patterns are also required to execute the automated condition monitoring tasks effectively.

You can tune multi-detector analytics and manage the model lifecycle by adjusting fusion configurations and performing parameter tuning. The framework supports model promotion from coldstart management through active deployment to ensure accurate condition monitoring.

The best way to manage coldstarts is through the framework's dedicated coldstart management feature within the batch run pipeline. This handles initial data ingestion and baseline model setup for industrial equipment before continuous multi-detector analytics take over.