diff --git a/README.md b/README.md
index da18514de81..3f67d0717af 100644
--- a/README.md
+++ b/README.md
@@ -6,6 +6,7 @@ Please make sure you have properly followed the LeRobot official installation (*
 
 [![watch the video](./src/lerobot/ufactory_usage/assets/inference-gello.png)](https://www.youtube.com/watch?v=wTiWLiHciT8)
 
+[![watch the video](./src/lerobot/ufactory_usage/assets/inference-pika.png)](https://youtu.be/IiyvewZh5OY)
 
 ## Acknowledgement:
 
diff --git a/src/lerobot/ufactory_usage/ReadMe.md b/src/lerobot/ufactory_usage/ReadMe.md
index 98fba3599cc..b3c65cff3b4 100644
--- a/src/lerobot/ufactory_usage/ReadMe.md
+++ b/src/lerobot/ufactory_usage/ReadMe.md
@@ -1,6 +1,7 @@
 # UFACTORY Data Collection Guide(Imitation Learning)
 
 ## ⚠️ Important Notice
+Using gello to collect dataset:
 - Once data collection starts, **the relative position between the robot arm and the camera (D435/D435i) must not be changed**.  
 - When making the inference, the position of the camera and the collection must be the same. Otherwise, the collected data will become invalid.  
 - In case of change, the entire process must be re-performed:
@@ -9,6 +10,10 @@
 Collect Dataset → Train the model → Perform Inference and Evaluation
 ```
 
+Using pika to collect dataset:
+- There are no specific requirements for the relative positions of the two pika stations and the robotic arm. However, **the pika needs to be recalibrated after the station moves**.
+- The positions of the pika stations during the collection and inference processes do not have to be the same.
+
 ## 1. Hardware Requirements
 
 ### Reference Hardware Setup
@@ -20,43 +25,48 @@ Collect Dataset → Train the model → Perform Inference and Evaluation
 - **Camera Mount**: Provided by UFACTORY (purchase or 3D print)
   - Purchase: [UFACTORY Camera Stand](https://www.ufactory.cc/product-page/ufactory-xarm-camera-stand/)
   - 3D Model: [Realsense_Camera_Stand.STEP](https://www.ufactory.cc/wp-content/uploads/2024/05/CameraStand_1300.zip)
-- **Teleoperation Device**: gello
+- **Teleoperation Device**: gello, [pika](https://global.agilex.ai/products/pika)
 
 ---
 
 ## 2. Software Environment Setup
 
-### 2.1 Install librealsense
+### 2.1 Using gello
+
+#### Install librealsense
 
 Reference: [librealsense Linux Installation Guide](https://github.com/realsenseai/librealsense/blob/master/doc/distribution_linux.md)
 
-#### Register Public Key
+##### Register Public Key
 ```bash
 sudo mkdir -p /etc/apt/keyrings
 curl -sSf https://librealsense.realsenseai.com/Debian/librealsense.pgp | sudo tee /etc/apt/keyrings/librealsense.pgp > /dev/null
 ```
 
-#### Install APT HTTPS Support
+##### Install APT HTTPS Support
 ```bash
 sudo apt-get install apt-transport-https
 ```
 
-#### Add the server to the list of repository
+##### Add the server to the list of repository
 ```bash
 echo "deb [signed-by=/etc/apt/keyrings/librealsenseai.gpg] https://librealsense.realsenseai.com/Debian/apt-repo `lsb_release -cs` main" | \
 sudo tee /etc/apt/sources.list.d/librealsense.list
 sudo apt-get update
 ```
 
-#### Install librealsense
+##### Install librealsense
 ```bash
 sudo apt-get install librealsense2-dkms
 sudo apt-get install librealsense2-utils
 ```
+### 2.2 Using pika
+Please refer to [pika tracking device calibration](https://github.com/xArm-Developer/ufactory_teleop/tree/main/pika_teleop).  
+First-time use of Pika Sense or when the base station position changes requires calibration.
 
----
 
-### 2.2 Create Python Virtual Environment
+## 3. Lerobot Environment Setup
+### 3.1 Create Python Virtual Environment
 
 #### Linux (Conda)
 ```bash
@@ -69,9 +79,8 @@ conda activate lerobot
 conda install ffmpeg -c conda-forge
 ```
 
----
 
-### 2.3 Install LeRobot and Dependencies
+### 3.2 Install LeRobot and Dependencies
 
 #### Install LeRobot
 ```bash
@@ -93,9 +102,9 @@ pip install pyrealsense2  # 2.56.5.9235
 pip install git+https://github.com/xArm-Developer/xArm-Python-SDK.git
 ```
 
-## 3. Data Collection
+## 4. Data Collection
 
-### 3.1 Teleoperation Test
+### 4.1 Teleoperation Test
 
 #### Using gello (xArm7)
 1. Modify the configuration file (example: `xarm7_gello_record_config.yaml`)
@@ -105,17 +114,35 @@ pip install git+https://github.com/xArm-Developer/xArm-Python-SDK.git
      - `start_joints`: Initial joint positions
    - **TeleoperatorConfig**
      - `start_joints`: Initial joint positions
-2. Move the gello joints to match the robot's initial pose
+2. Move the gello joints to match the robot's initial pose;
 3. Start the script:
 ```bash
 python uf_robot_teleop_test.py --config config/xarm7_gello_record_config.yaml
 ```
-4. Press **Enter** to begin teleoperation
-5. Press **ESC** to exit
+4. Press **Enter** to begin teleoperation;
+5. Press **ESC** to exit.
 
----
 
-### 3.2 Data Recording
+#### Using pika (xArm7)
+1. Modify the configuration file (example: `xarm7_pika_record_config.yaml`)
+   - **RobotConfig**
+     - `robot_ip`: IP address of the robot
+     - `robot_dof`: Degrees of freedom
+     - `start_joints`: Initial joint positions
+   - **TeleoperatorConfig**
+     - `start_joints`: Initial joint positions
+     - `gripper-port`: Pika gripper serial port
+     - `index_or_path`: Pika gripper camera index
+     - `port`: Pika sense serial port
+2. Move the pika sense to the initial position;
+3. Start the script:
+```bash
+python uf_robot_teleop_test.py --config config/xarm7_pika_record_config.yaml
+```
+1. Double click pika sense to start teleoperation;
+2. Press **ESC** to exit.
+
+### 4.2 Data Recording
 **Users may define their own data collection strategies. LeRobot only enforces the dataset format.**
 
 **Keyboard control** for data recording:  
@@ -134,24 +161,30 @@ python uf_robot_record.py --config config/xarm7_gello_record_config.yaml
 python uf_robot_record.py --config config/xarm7_mock_record_config.yaml
 ```
 
----
+#### Recording with pika(teleoperation)
+```bash
+python uf_robot_record.py --config config/xarm7_pika_record_config.yaml
+```
+
+### 4.3 Resume Data Recording
 
-### 3.3 Resume Data Recording
+#### Resume gello-based recording
 ```bash
-# Resume gello-based recording
 python uf_robot_record.py --config config/xarm7_gello_record_config.yaml --resume
 ```
-
+#### Resume random target recording
 ```bash
-# Resume random target recording
 python uf_robot_record.py --config config/xarm7_mock_record_config.yaml --resume
 ```
 
----
+#### Resume pika-based recording
+```bash
+python uf_robot_record.py --config config/xarm7_pika_record_config.yaml --resume
+```
 
-## 4. Training
+## 5. Training
 
-### 4.1 Initial Training(Use ACT)
+### 5.1 Initial Training(Use ACT)
 ```bash
 # Note:
 # repo_id must match the value used during data collection
@@ -168,7 +201,7 @@ python -m lerobot.scripts.lerobot_train \
   --steps=800000
 ```
 
-### 4.2 Resume Training
+### 5.2 Resume Training
 ```bash
 python -m lerobot.scripts.lerobot_train \
   --dataset.repo_id=ufactory/xarm7_record_datas \
@@ -184,7 +217,7 @@ python -m lerobot.scripts.lerobot_train \
   --config_path=outputs/train/xarm7_record_datas/checkpoints/last/pretrained_model/train_config.json
 ```
 
-## 5. Inference & Evaluation
+## 6. Inference & Evaluation
 
 ### Run with a Specified Model
 ```bash
@@ -192,7 +225,7 @@ python uf_robot_eval.py --config config/xarm7_gello_record_config.yaml \
   --policy.path=outputs/train/xarm7_record_datas/checkpoints/last/pretrained_model/
 ```
 
-## 6. Dataset Utilities
+## 7. Dataset Utilities
 
 ### Playback an Episode
 Example: view episode index **17**
@@ -222,25 +255,24 @@ aggregate_datasets(
 ```
 
 
-## 7. Important Notes
+## 8. Important Notes
 Users are expected to thoroughly study the codebase and configuration parameters.  
 The provided configurations are **not guaranteed to work for all scenarios** and must be adjusted based on actual hardware setups and task requirements.
 
 In particular, for **diffusion policies**, the default parameters in LeRobot are primarily designed for simulation and **are not optimized for real-world robots**.
 
 
-## 8. Dataset Examples (Reference Only)
+## 9. Dataset Examples (Reference Only)
 [Test datasets](https://drive.google.com/drive/folders/1Ms25rd2YYGdh3tHPEsTTMU-m1fE7uNYY?usp=sharing) used during development (not reusable):
 - **xarm7_act_20260119**: 60 successful single-attempt grasps recorded via gello
 - **xarm7_act_20260127**: 80 samples (added 20 failure-and-retry cases)
 - **xarm7_act_mock_20260126**: 60 one-shot grasps generated programmatically
+- **xarm7_pika_datas_aa_merge_0208_0209_2pi**: 150 samples grasps recorded via pika
 
 > These datasets are **for reference only**.  
-> They cannot be reused because the robot–camera calibration differs between users.
+> They cannot be reused because the robot–camera position differs between users.
 
-* Train using xarm7_act_20260127 for 500,000 iterations.
+* Train using xarm7_act_20260127 for 500,000 training steps.
   [![watch the video](/src/lerobot/ufactory_usage/assets/inference-gello.png)](https://www.youtube.com/watch?v=wTiWLiHciT8)
-* Train xarm7_act_mock_20260126 for 500,000 iterations.
-  --waiting...
-* Failure Cases
-  --waiting...
\ No newline at end of file
+* Train xarm7_pika_datas_aa_merge_0208_0209_2pi for 400,000 training steps.
+  [![watch the video](/src/lerobot/ufactory_usage/assets/inference-pika.png)](https://youtu.be/IiyvewZh5OY)
diff --git a/src/lerobot/ufactory_usage/ReadMe_cn.md b/src/lerobot/ufactory_usage/ReadMe_cn.md
index e2f193a5e37..6453917ac23 100644
--- a/src/lerobot/ufactory_usage/ReadMe_cn.md
+++ b/src/lerobot/ufactory_usage/ReadMe_cn.md
@@ -1,7 +1,9 @@
 # UFACTORY LeRobot 数据采集、训练与推理使用说明
 
-## ⚠️ 重要提示
+本文章仅提供数据采集方式，具体使用何种方式进行训练和推理客户可自行决定。
 
+## ⚠️ 重要提示
+对于gello数据采集方式：
 - **一旦开始数据采集，机械臂与摄像头（D435 / D435i）的相对位置必须保持不变**。
 - 推理时的摄像头位置必须与采集时相同。若机械臂或摄像头发生变化，**此前采集的数据将无效**。
 - 如发生变化，必须重新执行完整流程：
@@ -10,6 +12,11 @@
 重新采集数据 → 重新训练模型 → 重新推理运行
 ```
 
+对于pika数据采集方式：
+- **两个基站和机械臂相对位置没有要求，只需要保证采集时pika sense在基站范围内，但基站移动后需要重新校准**。
+- 采集和推理时基站位置可不相同。
+
+
 ## 1. 硬件要求
 
 ### 1.1 推荐硬件配置（示例脚本）
@@ -20,40 +27,47 @@
 - **摄像头支架**: UFACTORY 提供 (可购买或 3D 打印)
   - 购买链接: [UFACTORY 摄像头支架](https://www.ufactory.cc/product-page/ufactory-xarm-camera-stand/)
   - 3D 文件下载: [Realsense_Camera_Stand.STEP](https://www.ufactory.cc/wp-content/uploads/2024/05/CameraStand_1300.zip)
-- **遥操设备**： gello
+- **遥操设备**： gello, [pika](https://global.agilex.ai/products/pika)
 
-## 2. 软件环境配置
+## 2. 遥操设备环境配置
 
-### 2.1 安装librealsense库
+### 2.1 gello采集方式
+
+#### 安装librealsense库
 
 参考：[librealsense](https://github.com/realsenseai/librealsense/blob/master/doc/distribution_linux.md)
 
-#### 注册公钥
+##### 注册公钥
 ```bash
 sudo mkdir -p /etc/apt/keyrings
 curl -sSf https://librealsense.realsenseai.com/Debian/librealsense.pgp | sudo tee /etc/apt/keyrings/librealsense.pgp > /dev/null
 ```
 
-#### Https Support
+##### Https Support
 ```bash
 sudo apt-get install apt-transport-https
 ```
 
-#### 添加服务到仓库列表
+##### 添加服务到仓库列表
 ```bash
 echo "deb [signed-by=/etc/apt/keyrings/librealsenseai.gpg] https://librealsense.realsenseai.com/Debian/apt-repo `lsb_release -cs` main" | \
 sudo tee /etc/apt/sources.list.d/librealsense.list
 sudo apt-get update
 ```
 
-#### 安装librealsense库
+##### 安装librealsense库
 ```bash
 sudo apt-get install librealsense2-dkms
 sudo apt-get install librealsense2-utils
 ```
 
-### 2.2 创建Python虚拟环境
+### 2.2 pika采集方式
+请参考[pika安装和校准说明](https://github.com/xArm-Developer/ufactory_teleop/blob/main/pika_teleop/README_ZH.md)。  
+首次使用Pika Sense 或者**基站位置变动**需要校准。
 
+## 3. lerobot环境配置
+
+### 3.1 创建Python虚拟环境
 #### Linux(Using Conda)
 ```bash
 conda create -y -n lerobot python=3.10
@@ -65,7 +79,7 @@ conda activate lerobot
 conda install ffmpeg -c conda-forge
 ```
 
-### 2.3 安装lerobot和依赖
+### 3.2 安装lerobot和依赖
 #### 安装lerobot
 ```bash
 git clone https://github.com/xArm-Developer/lerobot.git
@@ -86,9 +100,9 @@ pip install pyrealsense2  # 2.56.5.9235
 pip install git+https://github.com/xArm-Developer/xArm-Python-SDK.git
 ```
 
-## 3. 数据采集
+## 4. 数据采集
 
-### 3.1 Teleop控制测试
+### 4.1 Teleop控制测试
 #### 使用gello(xArm7为例)
 1. 修改配置文件这里举例使用xarm7_gello_record_config.yaml
 * RobotConfig:
@@ -106,8 +120,28 @@ python python uf_robot_teleop_test.py --config config/xarm7_gello_record_config.
 4. 按下回车开始控制
 5. 按下'ESC'退出
 
+#### 使用pika(xArm7为例)
+1. 修改配置文件这里举例使用xarm7_pika_record_config.yaml
+* RobotConfig:
+  * robot_ip: 要控制的机械臂IP
+  * robot_dof: 机械臂轴数
+  * start_joints: 机械臂初始位置
+* TeleoperatorConfig:
+  * start_joints: 机械臂初始位置
+  * gripper_port：pika gripper的串口，可以不修改，会默认识别
+  * index_or_path：pika gripper上的鱼眼相机index
+  * port：pika sense的串口，可以不修改，会默认识别
 
-### 3.2 采集
+2. 将pika移动到操作的初始位置
+3. 启动脚本
+```bash
+python python uf_robot_teleop_test.py --config config/xarm7_pika_record_config.yaml
+```
+
+4. 双击pika gripper开始和关闭控制
+5. 按下'ESC'退出
+
+### 4.2 采集
 **用户可自定义采集方式，lerobot仅定义数据集的格式。**  
 
 以下举例使用gello采集或程序生成随机点进行采集：  
@@ -118,32 +152,36 @@ python python uf_robot_teleop_test.py --config config/xarm7_gello_record_config.
 ```bash
 python uf_robot_record.py --config config/xarm7_gello_record_config.yaml
 ```
-[![演示视频](assets/recording-with-gello.png)](https://www.bilibili.com/video/BV13jFjz2Eqq/?spm_id_from=333.1387.homepage.video_card.click)
-
 
 #### 使用程序生成随机目标点后, 程序自动生成数据来采集
 ```bash
 python uf_robot_record.py --config config/xarm7_mock_record_config.yaml
 ```
 
-[![演示视频](assets/recording-with-scripts.png)](https://www.bilibili.com/video/BV1tpf9BVEvp/?spm_id_from=333.1387.homepage.video_card.click)
-
+#### 使用pika采集(遥操)
+```bash
+python uf_robot_record.py --config config/xarm7_pika_record_config.yaml
+```
 
-### 3.3 恢复采集
+### 4.3 恢复采集
+#### gello恢复采集
 ```bash
-# gello恢复采集
 python uf_robot_record.py --config config/xarm7_gello_record_config.yaml --resume
 ```
 
+#### 程序生成点恢复采集
 ```bash
-# 程序随机生成点方式恢复采集
 python uf_robot_record.py --config config/xarm7_mock_record_config.yaml --resume
 ```
 
+#### pika恢复采集
+```bash
+python uf_robot_record.py --config config/xarm7_pika_record_config.yaml --resume
+```
     
-## 4. 训练
+## 5. 训练
 
-### 4.1 初次训练
+### 5.1 初次训练
 
 ```bash
 # 注意: repo_id就是采集时配置文件里面的repo_id
@@ -159,7 +197,7 @@ python -m lerobot.scripts.lerobot_train \
 --steps=800000
 ```
 
-### 4.2 恢复训练
+### 5.2 恢复训练
 ```bash
 python -m lerobot.scripts.lerobot_train \
 --dataset.repo_id=ufactory/xarm7_record_datas \
@@ -175,7 +213,7 @@ python -m lerobot.scripts.lerobot_train \
 --config_path=outputs/train/xarm7_record_datas/checkpoints/last/pretrained_model/train_config.json
 ```
 
-## 5. 推理
+## 6. 推理
 
 ### 指定模型进行推理
 ```bash
@@ -184,7 +222,7 @@ python uf_robot_eval.py \
 --policy.path=outputs/train/xarm7_record_datas/checkpoints/last/pretrained_model/
 ```
 
-## 6. 数据集工具
+## 7. 数据集工具
 提供一些数据集工具，方便对采集的数据集进行增删查操作。
 
 ### 查看某个索引的episode:
@@ -216,14 +254,16 @@ aggregate_datasets(
     )
 ```
 
-### 7.重要提示
+### 8.重要提示
 用户需要全面研究整个代码库，并了解相关的配置参数，因为代码中所写的配置并非适用于所有使用场景和设置，所以用户需要研究代码或相关理论，以获取相关知识，并自行进行修改和调整。特别是对于**扩散策略(diffusion policy)**，LeRobot 中的默认参数可能仅用于模拟，并未针对实际机器人场景进行优化。
 
-### 8. 数据集实例
+### 9. 数据集实例
 以下为开发测试时[采集的数据集](https://drive.google.com/drive/folders/1Ms25rd2YYGdh3tHPEsTTMU-m1fE7uNYY?usp=sharing)，**仅供参考，不可复用**，因为用户机械臂和摄像头位置和测试时不一致。
-* xarm7_act_20260119: gello手工录制一次性抓取成功的点，60个
-* xarm7_act_20260127: 在xarm7_act_20260119基础上增加了失败重抓的点20个，总共80个
-* xarm7_act_mock_20260126: 程序生成的60个一步到位的点
+* xarm7_act_20260119: gello手工录制一次性抓取成功的点(**关节运动**)，60组；
+* xarm7_act_20260127: 在xarm7_act_20260119基础上增加了失败重抓的点20组，共80组；
+* xarm7_act_mock_20260126: 程序生成的一步到位的点，60组；
+* xarm7_pika_datas_aa_merge_0208_0209_2pi：pika遥操采集的点(**笛卡尔-轴角**)，150组。
+  * 提示：使用笛卡尔运动时，无论使用RPY还是轴角采集时，数据Roll或Rx会存在不连续跳变问题，通过`±2π`可解决该问题，示例脚本中已处理。
 
 训练后效果：
 * 使用xarm7_act_20260127进行训练，训练50w次。
@@ -232,5 +272,8 @@ aggregate_datasets(
 * 使用xarm7_act_mock_20260126，训练50w次。
 [![演示视频](assets/inference-scripts.png)](https://www.bilibili.com/video/BV14pf9BVEox/?spm_id_from=333.1387.homepage.video_card.click)
 
+* 使用xarm7_pika_datas_aa_merge_0208_0209_2pi，训练40w次。
+[![演示视频](assets/inference-pika.png)](https://www.bilibili.com/video/BV16ccizHE2P/?spm_id_from=333.1387.homepage.video_card.click)
+
 * 失败案例
 [![演示视频](assets/failure-cases.png)](https://www.bilibili.com/video/BV1Waf9BTEMr/?spm_id_from=333.1387.homepage.video_card.click)
\ No newline at end of file
diff --git a/src/lerobot/ufactory_usage/assets/inference-gello.png b/src/lerobot/ufactory_usage/assets/inference-gello.png
index bb11faf7ac5..256184747b9 100644
Binary files a/src/lerobot/ufactory_usage/assets/inference-gello.png and b/src/lerobot/ufactory_usage/assets/inference-gello.png differ
diff --git a/src/lerobot/ufactory_usage/assets/inference-pika.png b/src/lerobot/ufactory_usage/assets/inference-pika.png
new file mode 100644
index 00000000000..83d72809d9c
Binary files /dev/null and b/src/lerobot/ufactory_usage/assets/inference-pika.png differ
diff --git a/src/lerobot/ufactory_usage/assets/recording-with-gello.png b/src/lerobot/ufactory_usage/assets/recording-with-gello.png
deleted file mode 100644
index 1baae5d1150..00000000000
Binary files a/src/lerobot/ufactory_usage/assets/recording-with-gello.png and /dev/null differ
diff --git a/src/lerobot/ufactory_usage/assets/recording-with-scripts.png b/src/lerobot/ufactory_usage/assets/recording-with-scripts.png
deleted file mode 100644
index 493a10e8b28..00000000000
Binary files a/src/lerobot/ufactory_usage/assets/recording-with-scripts.png and /dev/null differ