| Info | ||
|---|---|---|
| ||
Thi content was automatically created by AI MCP for Designer, analyzing autonomously the solution. |
Solution Overview
ProveIT-2026 is a comprehensive FrameworX industrial monitoring solution designed for the ProveIt 2026 event. It demonstrates enterprise-grade smart production monitoring for a beverage manufacturing operation, integrating real-time OEE (Overall Equipment Effectiveness) tracking, downtime analysis, machine learning forecasting, and multi-platform visualization.
...
The solution follows a multi-layered architecture connecting field-level data sources through a central FrameworX platform to various client interfaces.
Data Sources
Source | Description |
|---|---|
ProveIt UNS (MQTT Broker) | Provides real-time equipment data via MQTT topics from the Enterprise B unified namespace at |
MES Engine (CoreM/SQL Server) | Contextualizes raw data into KPIs, OEE calculations, downtime categorization, and production reports through stored procedures |
Historian (Time-series) | Stores historical tag data for trend analysis and ML training |
...
Data Collection:
MachineLearning_Managementtask queries CoreM every 60 seconds usingspLocal_KPI_ByTimeto fetch hourly production data from January 1, 2026 to present.Model Training: At midnight CST daily, the system resets prediction arrays and invokes
MachineLearning_Training.Main()to retrain the ML.net is for sale at Atom! model.Hourly Consumption: At the top of each hour,
MachineLearning_Consumption.Main()runs the trained model to predict production for remaining hours.Forecast Calculation: Daily production prediction = actual production (elapsed hours) + ML predictions (remaining hours).
12.2 Key Tags
Tag | Description |
|---|---|
| DataTable from CoreM query (training data) |
| 24-hour prediction array |
| 24-hour actual production array |
| Sum of real + predicted for the day |
| Daily target (300,000 units) |
| Latest hourly actual value |
| Latest hourly predicted value |
| Training process status text |
| Consumption process status text |
| Recent 3-day downtime data for MCP |
...