✨ New Features
1. Advanced Heat Map Visualization
The Heat Map object has been significantly enhanced to support real-world analytical use cases. It now includes native input connectors, multiple data entry modes (static tables, DataFrame Creator inputs, and live connectors), and improved color handling with automatic color scheme selection. Labels, titles, and scale presentation have been refined to support clear interpretation and direct use in reports and presentations.
2. Automated Historic Data Preloading for Delayed Inputs
SharpBot can now automatically populate delayed input queues using historical data from PI servers. This enables models with long input delays to start producing predictions immediately, without waiting hours for buffers to fill, improving usability for real-time deployment and commissioning.
3. Resizable Correlation Tables in DataFrame Creator
Correlation tables (Pearson and Spearman) in the DataFrame Creator are now resizable with the dialog window. This makes it much easier to analyze large datasets, review correlation patterns, and avoid unnecessary exports to external tools.
🐞 Bug Fixes
1. Reliable Time Alignment in DataFrame Creator
Resolved an issue where time-series variables were merged without proper resampling. DataFrame Creator now consistently aligns all variables to a common time reference, preventing silent data corruption and improving analytical accuracy.
2. Correct Column Type Detection on Import
Improved auto-detection logic to prevent numeric columns from being incorrectly interpreted as date fields during first-time file imports. Column headers and values are now reliably preserved as intended.
3. Accurate Handling of Fractional Time Ranges
Resolved display inconsistencies when specifying fractional days for PI data retrieval. Decimal values are now clearly shown and consistently interpreted, reducing confusion during configuration.
4. Optimizer Responsiveness to Configuration Changes
Resolved issues where changes to constraints or initial conditions were not applied correctly. The Optimizer now initializes from live input values and immediately responds to constraint updates, ensuring consistent optimization behavior.
5. Target Seeking Restored in Parameter Estimation
Restored target-seeking functionality for both single-input and multi-input parameter estimation workflows, enabling stable convergence and more predictable solution paths.
6. Stable Log File Rotation
Improved log file handling so new daily log files are created automatically when SharpBot remains open overnight, ensuring clear diagnostics and easier troubleshooting.
🎨 UI & Usability Improvements
1. Safer, Neutral Integration Icons
Replaced trademarked third-party logos with neutral, purpose-built icons for PI, OPC, Excel, and related interface blocks. This ensures consistent visual language while avoiding unnecessary legal exposure.
2. Improved Default Block Initialization
Value get/set blocks now initialize with safer default references, reducing the risk of accidental connections and improving first-use reliability.
3. Clearer Labels and Scaling in Visual Tools
Visualization components now present titles, scales, labels, and gradients more consistently, improving clarity when analyzing or exporting visual results.
⚙️ Performance & Stability Improvements
1. More Predictable Optimizer Initialization
The Optimizer now reliably initializes at the plant’s current operating state by reading live input values, improving convergence speed and reducing manual setup.
2. General Stability Enhancements
Multiple internal fixes across data handling, visualization rendering, and block initialization contribute to smoother operation and greater confidence in continuous-use scenarios.
🔌 Connectivity & Integration Enhancements
1. Correct OPC Tag List Export
Tag list exports from the OPC Client now correctly reflect only the tags used in the model and include meaningful metadata such as data type, scaling, quality, and read/write status.
2. Consistent PI Server Naming
Updated terminology across SharpBot to consistently refer to “PI Server,” providing clearer and more vendor-neutral integration labeling.
📘 Documentation & Examples
1. Updated Examples for Optimization and Visualization
Example models have been refreshed to reflect the latest Optimizer behavior and visualization enhancements, helping users adopt new capabilities more quickly and confidently.