AugeLab Studio - What's New?
Stay up-to-date with the latest improvements and features in AugeLab Studio.
Version 3.11.1 (September 2026)
Bug Fixes & Improvements
- Linux: Fixed a startup crash after updating.
- Node Editor: Fixed context menu placement and styling.
- Windows Setup: Improved graphics compatibility.
Version 3.11.0 (September 2026)
Highlights
- New AI Agent: Updated AI agents across all versions.
- Training Images: Compare image variations with the original and reuse your saved settings.
- Industrial Cameras: Easier setup for Hikrobot and Irayple cameras.
- Dataset Overview: See label totals and category counts in a clearer chart.
- Windows Setup: Automatically selects the right AI package for compatible NVIDIA graphics cards.
Bug Fixes & Improvements
- Cameras: More reliable connections, clearer error messages, and corrected Irayple images on Linux.
- Linux: Improved image and video support.
- AI Training: Clearer warnings and setup help when missing graphics-card support would slow down training.
- Interface: More reliable windows, notifications, and progress displays.
- Auto-Complete: Fixed suggestions in Data Store, Tag, and similar fields.
- Irayple Cameras: Find cameras directly by IP address and receive images across supported pixel formats.
- FTP Download: Automatically checks for changed remote timestamps and missing local files, with Success set to True for one step after each completed download.
- FTP Upload: Completion results remain available until another transfer starts, including when the block is disabled during a transfer.
- Offline Startup: Reduced waits for unavailable online services when falling back to offline licensing.
- Custom Blocks: Locally installed custom blocks load without Marketplace authorization, including in offline mode.
- Linux Installation: Skips reinstalling system packages that are already present on Linux and Linux ARM64.
- Scenario Recovery: Blocks that encounter execution errors clear their outputs and can retry on the next step without leaving other blocks waiting.
- Command-Line Runs: A closed console output pipe no longer causes logging errors to interrupt scenario execution.
Version 3.10.0 (August 2026)
Highlights
- AugeLab AI: A next-generation AI platform upgrade across training and inference, with a more complete production experience.
- Smart Training: AugeLab will check your AI training progress and will stop earlier or extend your training as needed.
- Image Classification: Faster dataset onboarding, guided training flows, and smoother use of classification models in Studio.
- Anomaly Detection: A cleaner pipeline for anomaly scoring, detection output, and daily monitoring workflows.
- PADIM and AUGE: New dedicated training/validation paths for PADIM and AUGE use cases.
- FTP & FTPS Transfers: New blocks for secure FTP connect, upload, and download to simplify data movement.
- Color Consistency: Better color-space control so visual results stay stable from capture to decision.
Bug Fixes & Improvements
- AI Training Stability: Safer long-run behavior through checkpointed training workflows and stronger recovery.
- AI Training Feedback: More useful dashboard views, progress signals, and reliability during extended runs.
- Annotation Reliability: Dataset switching now restores context more predictably and avoids stale class/session states.
- Annotation Conversion Safety: COCO import and annotation conversion now use safer path handling and cleaner folder transitions.
- Object Detection Compatibility: Legacy Object Detection output behavior is preserved for existing scenes.
- Camera Throughput: Industrial camera pipelines run with lower overhead and better production reliability.
Version 3.9.3 (July 2026)
Highlights
- Allen-Bradley PLC Support: Added blocks for connecting to Allen-Bradley controllers and reading or writing PLC tags.
- Save Video: Added a simpler video-recording block that saves timestamped H.264 MP4 files in the background with selectable quality.
- Industrial Cameras: Basler, Hikrobot, and Irayple cameras now share a more consistent setup and capture experience.
- YOLO Training Settings: The training window now remembers your dataset, classes, model, weights, and advanced settings between sessions.
Bug Fixes & Improvements
- YOLO Training: Improved dataset checks, model configuration, training progress, completion handling, anchor calculation, and model export reliability. Mosaic augmentation is now disabled by default.
- Scope Block: Improved chart updates and data handling for more reliable live plots.
- Multi-Port Switch: Improved handling of empty values and preserved the previous behavior through the legacy block.
- Irayple Cameras: Improved camera-library loading on Linux.
- Camera Performance: Increased frame rates for supported industrial cameras.
Version 3.9.2 (July 2026)
Bug Fixes & Improvements
- AI Assistant: Improved the reliability of chat responses and activity updates while the assistant works.
- HMI Backgrounds: Background areas now behave more reliably when working with HMI scenes.
- YOLOv4 Training: Fixed a configuration issue that could prevent YOLOv4-tiny training from starting correctly.
- Object Detection Setup: Updated the required AI components for more reliable object-detection workflows.
- GPU compatibility: Included improved GPU compatibility for older devices.
Version 3.9.1 (July 2026)
Bug Fixes & Improvements
- AI Block Fix: Fixed installation issues that could prevent Studio from setting up the required Object Detection packages correctly.
Version 3.9.0 (July 2026)
Highlights
- Open .pmod Files Directly: Open Studio scenarios from your file manager or operating system file associations more smoothly.
- More Resilient AI Assistant: Long assistant actions can recover better after interruptions and show clearer progress while they work.
- Improved Scene Assistance: Assistant-guided scene editing, scene checks, plugin creation, and HMI setup now handle more real projects with clearer results.
- Object Detection & Training: Object detection and YOLO workflows now choose the right runtime more reliably across CPU, GPU, and Jetson setups.
- Anomaly Detection Example: Added an updated custom block example for training on normal images and producing heat maps and masks.
Bug Fixes & Improvements
- Startup & Support Logs: Startup failures are captured more consistently and included when exporting logs for support.
- Scene Cleanup Stability: Closing, reloading, and automated scene runs now clean up graphics more reliably, reducing crash risk.
- ROI Editing: Circular ROI mouse detection is more reliable, and cleared polygon ROIs no longer cause execution errors.
- Subsystem Display: Fixed a display issue that could appear after switching between subsystems.
- CUDA Setup: CUDA discovery and fallback behavior is more reliable, especially on Jetson and Linux systems.
- Installer & File Associations: Installer handling for opening and removing Studio file associations is more dependable.
Version 3.8.0 (June 2026)
Highlights
- AI Assistant History: Reopen previous assistant conversations and continue them later without losing context.
- Local MCP Server: Connect approved local AI clients to Studio for scene inspection, node discovery, graph editing, logs, documentation, and scene execution.
- AI Documentation Access: The assistant can now search bundled Studio guidance and AugeLab documentation more directly.
- HMI Focus Window: Parametric image widgets can open in a focused view for easier inspection during HMI workflows.
Bug Fixes & Improvements
- Assistant Reliability: Tool errors are clearer, scene operations report real results, and one-step scene runs now return after completion instead of leaving clients waiting.
- MCP Configuration: Added a settings screen and automatic port fallback so local MCP setup is easier when the default port is busy.
- Recent Scenarios: Recent files are now ordered by last opened time, and session state is saved more frequently.
- Safer Closing: Accidentally closing Studio now respects confirmation dialogs more reliably.
- Updater Feedback: Update problems now show clearer messages so users know what went wrong.
- Startup Stability: Fixed startup cases where splash or startup dialogs could get stuck.
- Subsystem Blocks: Improved socket type handling for subsystem inputs and outputs.
- Linux Camera Support: Improved Irayple camera SDK detection and loading on Linux systems.
Version 3.7.1 (June 2026)
Highlights
- Easier Subsystem Wiring: Create subsystem inputs and outputs directly from the main scene by hovering or dragging connections onto subsystem blocks.
- New Alignment Block: Align scene images using reference ROIs with translation-only or translation-plus-angle correction.
- Smarter Reference Entry: Data Store and Tag blocks now offer auto-complete.
Bug Fixes & Improvements
- Windows Installer: Startup is more reliable, with better compatibility handling and bundled runtime support.
- Startup Flow: Fixed cases where the splash screen could stay visible while startup dialogs were open.
- AI Chat Window: Closing the AI chat no longer freezes the application.
- Basler Camera Handling: Disconnected Basler cameras no longer keep reconnecting and now release more cleanly.
- Linux Window Controls: Custom title bar icons now display more consistently on Linux systems.
- License Guidance: License and activation problems now show clearer messages to help you recover faster.
Version 3.7.0 (June 2026)
Highlights
- New Block - Image ROI Circle: Create a circular ROI directly on the image preview, then move or resize it with interactive handles.
- Refreshed ROI Blocks: Perspective Transform, Image ROI Select Multi, Image ROI Polygon, and Image ROI Polygon Multi now support more practical on-image editing.
- Broader Linux Installation Support: Linux installers now support more distributions, including systems using apt, dnf, and pacman.
Bug Fixes & Improvements
- Linux ARM Support: Improved UI dependency handling and native library compatibility on ARM-based Linux systems.
- Linux Desktop Launchers: Installer-created shortcuts are more consistent for UI installations.
- Siemens S7 Communication: Improved read and write reliability for strings, booleans, and shared PLC connections.
- Perspective Transform: You can now select four points in any order and drag points to adjust the transform.
- Image ROI Select Multi: Multiple rectangles can now be created, moved, resized, selected, undone, or deleted more easily.
Version 3.6.1 (April 2026)
Highlights
- Improved Installer Experience: Windows installation now shows clearer progress, better logs, and a more reliable installer window.
- Linux Update Support: Studio can now launch Linux update installers from the update flow.
- Safer AI Bundle Installation: Studio now checks for NVIDIA graphics hardware before installing the AI Bundle on Windows and gives a clear warning before continuing.
Bug Fixes & Improvements
- Custom Blocks: Renaming a custom block now preserves its sockets more reliably.
- Divide Images Block: Fixed output scaling so divided images display with the expected brightness.
- Block List Stability: Fixed a crash that could happen when sorting blocks with mixed internal identifiers.
- Irayple Camera Support: Studio now handles missing Irayple SDK paths more gracefully.
- CUDA/OpenCV Setup: Improved runtime setup for GPU and OpenCV components during installation.
- Module Downloader: Simplified package installation behavior for more predictable add-on downloads.
Version 3.6.0 (March 2026)
Bug Fixes & Improvements
- Annotation Converter: Improved conversion reliability across supported annotation formats.
- YOLO Training: Fixed augmented images affecting validation splits, so training results are more trustworthy.
- Custom Block Search: Scene search now finds custom blocks more reliably.
- Draw Rectangle: Fixed status-based rectangle colors so overlays now display the correct pass/fail color.
- D-FINE Detection: Improved runtime stability for D-FINE based AI workflows.
- HMI History Viewer: Improved layout and scene-data viewing for a clearer run-history experience.
- Standalone HMI Export: Improved launcher/export reliability so standalone HMI packages start more consistently.
- List Operations: Fixed default socket setup when creating the node.
- Open Folder Actions: Folder shortcuts now work more reliably on both Linux and Windows.
- Irayple Camera Support: Fixed an import issue that could prevent the camera module from loading correctly.
- Installer & AI Modules: Linux and Windows install flows now handle optional AI packages more safely and predictably.
Version 3.5.1 (February 2026)
Highlights
- Parallel Execution Stability: Improved reliability and predictability for complex scenes running under heavy load.
- OCR Reliability: Reduced random OCR failures in parallel workflows by improving how OCR inference is handled per block.
- Standalone HMI Improvements: Added stronger model compatibility checks with clearer guidance when HMI scene setup is incomplete.
- (BETA) Linux Deployment Support: Expanded Linux installer and packaging improvements across supported Linux targets.
Bug Fixes & Improvements
- Video Recording (Linux): Added safe fallbacks for keyboard/codec limitations with clearer in-app warnings.
- USB Camera Selection: Fixed external USB camera index handling so the intended device is selected more consistently.
- Image Memory Block: Fixed image persistence so saved images restore correctly after reload.
- Toolbar Usability: Toolbar tooltips now show keyboard shortcuts for faster workflow discovery.
- Camera Scanner Stability: Improved startup behavior when optional camera SDKs are unavailable.
Version 3.5.0 (January 2026)
Highlights
- Parallel Execution Mode: Introduced a new parallel execution mode for faster processing. You can now toggle this option in the settings.
- AI Infrastructure Upgrade: Updated to PyTorch 2.10 and Transformers 5.0, providing access to the latest AI features and improved model performance.
- New AI Node - U2Net Segmentation: Added a new AI node for high-quality background removal and object segmentation.
- Video Recording Enhancements: The
record_video block now supports high-compression options and exports videos in web-compatible formats.
Bug Fixes & Improvements
- Module Downloader: Enhanced the AI bundle installation process with better support for CUDA 12.8 and more robust dependency management.
- Text Rendering Improvements: Added background options and a wider range of colors to the
text_on_image block.
- Stability Fixes: Fixed a crash when closing the node editor and resolved node dragging issues after stopping a run.
- Histogram Analysis: Fixed peak detection and output issues in the
histogram_on_line block.
- Improved accuracy and efficiency of the performance analysis tools.
Version 3.4.6 (January 2026)
Highlights
- Improved Designer Experience: More robust custom block reloading that preserves scene stability and connections.
- Auto-Annotation Enhancements: Improved color management for automated annotation tasks.
- AI Training: Refined UI for Object Detection training window.
Bug Fixes & Improvements
- Enhanced file navigation in the Designer Window with new Refresh and Open Folder shortcuts.
- Integrated license-based limits for AI Assistant requests.
- Optimized internal imports and code structure for better performance.
- Fixed node replacement logic when updating existing custom blocks.
Version 3.4.5 (December 2025)
Highlights
- Linux compatibility improvements with native and docker installations.
Version 3.4.4 (December 2025)
Highlights
- AugeLab Studio now features an online installer and updater.
Version 3.4.3 (December 2025)
Highlights
- AI training performance update.
Version 3.4.2 (December 2025)
Highlights
- Package manager locks add-on installs to your current Studio version to avoid surprise upgrades.
Bug Fixes
- AI training sessions resume correctly after GPU and OpenCV service adjustments.
- Object detection training window launches without layout glitches.
- PaddleOCR modules install with the right CUDA build again.
Version 3.4.1 (December 2025)
✨ Highlights
- AI Assistant Everywhere: Ask the AI assistant about blocks, logs, and annotations with a simple right-click.
- Enhanced AI Chat: New modern interface with streaming responses and better interaction.
- Basler & Hikvision Cameras: Full support for industrial camera integration.
- Auto-Restart Dashboard: Monitor and configure system restart behavior with the new interface.
- Improved Performance: Faster block configuration and smoother overall experience.
Bug Fixes & Improvements
- Fixed MQTT reconnection issues.
- Improved Modbus client connection stability.
- Enhanced SQL database operations to prevent freezing.
- Fixed chart display issues (Scope block).
- Resolved image annotation rendering problems.
- Fixed "Open Recent" menu functionality.
- Added "Zip Logs" feature in Help menu for easier support.
- Improved notification system across all windows.
- Better marketplace error handling.
- Enhanced YOLO training performance.
Version 3.2.2 (October 2025)
Flash Fixes (AI Training included)
- AI training is not functional on Windows when the selected language uses a comma (,) as the decimal separator.
- Performance improvement on mAP calculation.
Version 3.2.1 (October 2025)
Flash Fixes
- Module downloader not opening.
Version 3.2.0 (October 2025)
Highlights
- New Blocks: Background removal, Background segmentation, Depth anything, Dfine model, Easy ocr, Match anything.
- Auto save feature.
- Input change detection improvement for SQL blocks.
- IP Camera and Youtube stream support is improved.
- Better ONVIF authentication support.
- More informative splash screen.
Bug Fixes
- Auto restart mode hardened.
- IP Camera fix on freeze with Windows UI.
- Object detection training graph fix on showing 0 values for loss and mAP.
- Auto-annotation saving issue.
- Augmentation bug-fix on some files.
- Output type for some blocks.
- Copy paste of some subsystems causing summary to be incorrect.
- Class frequency analysis without a loaded class file.
Version 3.1.3 (August 2025)
Flash Fix
- Fixed a bug where block input and output names are displayed incorrectly.
Version 3.1.2 (August 2025)
Highlights
- New training graph for object detection training.
- New block: Line Geometry
Bug Fixes & Improvements
- Performance improvement for designer window when copy-pasting huge text chunks.
- Bug fix on function blocks snapping to a random position when dragging.
- Scope block bug fix.
- Image translate block bug fix.
- Bug fix for socket names not displaying.
Version 3.1.0 (June 2025)
Highlights
- Modern Welcome Page: A friendlier, more informative welcome experience.
- Headless Server Dashboard: Remotely access and monitor your headless instance from anywhere.
- Custom Block Controls → Input Sockets: Instantly convert controls by right clicking into input sockets for flexible block design.
- AI Training Upgrades: Advanced live augmentation and fine-grained training control.
Bug Fixes & Improvements
- Clear error message when deleting blocks with connections.
- OPC UA reconnection is now more robust.
- Annotation file saving issue resolved.
- Augmentation bug fixes for improved reliability.
Version 3.0.5
Bug Fixes & Stability Improvements
- New marketplace explorer.
- MQTT TLS block have been implemented with beta version.
- Dataset augmentation fixes.
- AI training early finish fix.
Version 3.0.0
Exciting New Features!
- New Python Engine: Migrated to Python 3.12 for enhanced performance and updated package management.
- Performance Boost: Experience up to 30% performance improvements in various scenarios.
- Enhanced Custom Block Assistant: Improved user experience and functionality.
- Streamlined AI Training: More robust and efficient AI training workflows.
- Automatic Annotation: Introducing automatic annotation capabilities for faster data preparation.
Version 2.5.2
Bug Fixes & Stability Improvements
- Siemens S7 connection block reconnection issue resolved.
- YOLO training algorithm updated for improved accuracy.
- MQTT Reconnection stability enhanced.
Version 2.5
Key Updates
- Designer & UI Enhancements: Turkish character support added to Designer Window and improved image handling for Turkish characters.
- System & Command-Line Tool Updates: Modbus client connection stability improved.
- Image Processing Block Improvements: Multiple new and enhanced image processing blocks including GrabCut, FloodFill, Image Translation, Background Subtraction, and more.
- Layout & Widget Stability: Layout deserialization and widget conversion within layouts fixed.
- Logging & Error Handling Improvements: Enhanced image write logging, auto-restart bug fixes, and scene closing stability.
- Scene & Augmentation Updates: Augmentation "Select All", Subsystem switch enhancements, Rotation interpolation, Scene window closing fixes, and Cycle Time feature.
- Asynchronous Operations & Time Function Fixes: Async Siemens write race condition resolved.
- AI Training Window Enhancements: Auto GPU/RAM calculation, Batch Size adjustment, and Anchor calculation implemented.
- General Fixes & Improvements: Dragging ROI error fix, Mask application improvements, Weighted addition enhancements, Date-time naming and millisecond support, Simplified Image Write.
Version 2.2.2
New Fast AI Models & UI Updates
- New Fast AI Models (CPU Compatible): Micro and nano object detection models for CPU deployment.
- Custom Object Detection Block (CPU): New block for CPU-based object detection without GPU requirement.
- Enhanced HMI Window Block: Improved functionality and user experience.
- REST API Blocks (Asynchronous): REST API blocks now operate asynchronously.
- List Operations & Subsystem Loop Updates: Pairing lists with List Operations block and loop block improvements.
Version 2.2.1
UI/UX Improvements & Bug Fixes
- Colorful Connections: Visual enhancement with colorful connections in the designer.
- Enhanced HMI Window Block: Further improvements to HMI Window Block functionality.
- Bug Fixes: Addressed function block resizing, connection view issues with HMI window, Data Type Converter batching bug, and copy-paste order mixing.
Thank you for using AugeLab Studio!