Release Notes#
Click each tab to learn about the new and updated features in each release of Intel® Embodied Intelligence SDK.
Embodied Intelligence SDK v25.36 enhances model optimization capabilities with OpenVINO™ and provides typical workflows and examples, including Diffusion Policy (DP), Robotic Diffusion Transformer (RDT), Improved 3D Diffusion Policy (IDP3), Visual Servoing (CNS) and LLM Robotic Demo. This release also updated Real-time optimized BKC to improve on AI/Control performance, and supporting on Intel Arc B-Series Graphic card (B570).
New Features:
Updated Real-time optimization BKC, including BIOS and runtime optimization, balancing performance with AI & Control consolidation.
Added support for Intel Arc B-Series (Battlemage) Graphics card (B570).
Fixed deadlock issue when reading i915 perf event in Preempt-RT kernel.
New EtherCAT Master stack features supporting user-space EtherCAT Master and multiple EtherCAT masters.
Added Diffusion Policy pipeline with OpenVINO optimization.
Added Robotics Diffusion Transformer (RDT) pipeline with OpenVINO optimization.
Added Improved 3D Diffusion Policy (IDP3) model with OpenVINO optimization.
Added Visual Servoing (CNS) model with OpenVINO optimization.
Provided new tutorials for typical AI model optimization with OpenVINO.
ACRN initial enablement on ARL platform.
Added new Dockerfile to build containerized RDT pipeline.
Known Issues and Limitations
ACRN feature and performance
iGPU performance degradation observed when using passthrough iGPU to VM on ACRN.
Display becomes unresponsive in VMs when running concurrent AI workloads with iGPU SR-IOV enabled on ACRN.
The following model algorithms were added and optimized by Intel® OpenVINO™:
Algorithm |
Description |
|---|---|
Qwen2.5VL |
|
Whisper |
|
FunASR (Automatic speech recognition) |
Refer to the FunASR Setup in LLM Robotics sample pipeline |
A graph neural network-based solution for image servo utilizing explicit keypoints correspondence obtained from any detector-based feature matching methods |
|
A visuomotor policy learning model in the field of robotic visuomotor policy learning, which represents policies as conditional denoising diffusion processes |
|
A diffusion policy model enhancing capabilities for 3D robotic manipulation tasks |
|
A diffusion-based foundation model for robotic manipulation |
The following pipelines were added:
Pipeline Name |
Description |
|---|---|
An innovative method for generating robot actions by conceptualizing visuomotor policy learning as a conditional denoising diffusion process |
|
A RDT pipeline provided for evaluating the VLA model on the simulation task |
|
A code generation demo for robotics, interacting with a chatbot utilizing AI technologies such as large language models (Phi-4) and computer vision (SAM, CLIP) |
Embodied Intelligence SDK v25.15 provides necessary software framework, libraries, tools, BKC, tutorials and example codes to facilitate embodied intelligence solution development on Intel® Core Ultra Series 2 processors (Arrow Lake-H), It provides Intel Linux LTS kernel v6.12.8 with Preempt-RT, and supports for Canonical® Ubuntu® 22.04, introduces initial support for ROS2 Humble. It supports many models optimization with OpenVINO™, and provides typical workflows and examples including ACT manipulation, ORB-SLAM3, etc.
New Features:
Provided Linux* OS 6.12.8 BSP with Preempt-RT
Provided Real-time optimization BKC
Optimized IgH EtherCAT master with Linux* kernel v6.12
Added ACT manipulation pipeline with OpenVINO™/Intel® Extension for PyTorch optimization
Added ORB-SLAM3 pipeline focuses on real-time simultaneous localization and mapping
Provided typical AI models optimization tutorials with OpenVINO™
Known Issues and Limitations
There is a known deadlock risk and limitation to use
intel_gpu_topto read i915 perf event in Preempt-RT kernel, it will be fixed with next release.
The following model algorithms were optimized by Intel® OpenVINO™:
Algorithm |
Description |
|---|---|
CNN based object detection |
|
CNN based object detection |
|
CNN based object detection |
|
Transformer based segmentation |
|
Extend SAM to video segmentation and object tracking with cross attention to memory |
|
Lightweight substitute to SAM |
|
Lightweight substitute to SAM (Same model architecture with SAM. Can refer to OpenVINO SAM tutorials for model export and application) |
|
CNN based segmentation and diffusion model |
|
Transformer based object detection |
|
Transformer based object detection |
|
Transformer based image classification |
|
An end-to-end imitation learning model designed for fine manipulation tasks in robotics |
|
A self-supervised framework for interest point detection and description in images, suitable for a large number of multiple-view geometry problems in computer vision |
|
A model designed for efficient and accurate feature matching in computer vision tasks |
|
Obtaining a BEV perception is to gain a comprehensive understanding of the spatial layout and relationships between objects in a scene |
|
A powerful tool that leverages deep learning to infer 3D information from 2D images |
The following pipelines were added:
Pipeline Name |
Description |
|---|---|
Imitation learning pipeline using Action Chunking with Transformers(ACT) algorithm to train and evaluate in simulator or real robot environment with Intel optimization |
|
One of popular real-time feature-based SLAM libraries able to perform Visual, Visual-Inertial and Multi-Map SLAM with monocular, stereo and RGB-D cameras, using pin-hole and fisheye lens models |