DOCSEARCHbyAlgolia
Submit Docs

SparklingRobo Documentation

website·Indexed 30 days ago

https://docs.sparklingrobo.com

Documentation for Synria robotic systems, including the Alicia-M and Alicia-D series robotic arms, Gloria-D SDK, and Synria Desk. Covers PySDKs, ROS 2 Humble integration, teleoperation, embodied intelligence workflows (training, inference, and dataset management), and hardware components like the Synria C10 camera and tactile skins.

Tokens
144K
Snippets
645
Records
1.1K
Agent score
49%

What's inside SparklingRobo

  1. Overview of the Yunxi Gloria-M Parallel Two-Finger Motor Gripper

    The Yunxi Gloria-M Parallel Two-Finger Motor Gripper (v1.0.0) is a servo-motor-driven intelligent gripper designed for high-precision, programmable end-effector tasks in desktop robots and laboratory environments. It is compatible with all robotic arms and linear platforms and is intended for use in scientific research, teaching demonstrations, semi-automatic assembly, and laboratory automation. Key capabilities include sub-millimeter repeatable positioning, an open data protocol for secondary development, and a compatible ecosystem for expanding fingertips with vision assistance, tactile extensions, or compliant force control.

  2. Overview of Gloria-M Gripper Control

    The Gloria-M gripper control interface allows developers to connect to the gripper, monitor motor feedback, switch between PV and MIT control modes, adjust opening/closing positions and motion parameters, read internal parameters, and scan for CAN IDs. The interface includes a 3D model preview that synchronizes with the gripper's current opening degree (range 0-1000) and changes color based on the current torque to indicate gripping force.

  3. Overview of the Alicia-D Teleoperation Robotic Arm Series

    The Alicia-D Teleoperation Robotic Arm Series is an embodied-intelligence platform for AI education and industrial use. It uses a leader-follower architecture where the motion of a handheld leader arm is mapped to a follower arm via teleoperation coordination algorithms. The system is designed for robotic-arm control, imitation learning, and end-to-end large-model (VLA) research. It is fully open source and integrates with the Synria SparkMind Embodied-Intelligence Algorithm Platform for vision, control, and mechanical system integration.

  4. Overview of Alicia-DL Leader Arm capabilities

    The Alicia-DL Leader Arm series features a standard six-degree-of-freedom (6-DOF) design. It supports optional configuration packages for Aloha and UR architectures. Using integrated teleoperation collaborative algorithms, the leader arm's movements can be precisely mapped to various robotic arm configurations, including the proprietary Lingdong (灵动) operating arm, as well as arms from Aloha, PiPER (松灵), UR, and Flexiv (法奥).

  5. Overview of Alicia-D 6D Grasping (Eye-on-Hand)

    The alicia_d_6d_grasp module provides 6D grasping functionality for the Alicia-D robot arm by combining a depth camera with the GraspGen algorithm to grasp objects in arbitrary poses.

    Supported Cameras:

    • Intel Realsense D405 (Recommended): Supports colored point clouds.
    • Orbbec Gemini 335: Does not support colored point clouds, but grasping performance is unaffected.

  6. Overview of the Lingxi Gloria-D Standard Edition Gripper

    The Lingxi Gloria-D Standard Edition is a Parallel Two-Finger Servo Gripper designed for desktop robotic arms and linear platforms. It features a high-precision servo drive unit that supports millimeter-level repeatable positioning control. Due to its open communication protocol and simple electrical interface, it is suitable for teaching demonstrations, laboratory pick-and-place tasks, and lightweight automation.

  7. Overview of Synria Alicia-M Leader-Follower Teleoperation Series

    The Synria Alicia-M Leader-Follower Teleoperation Series is designed for light industrial applications, advanced research, and embodied intelligence data collection. It maps motions from a teaching arm (leader) to a follower arm for teleoperation, teaching data collection, imitation learning, and policy verification.

  8. Overview of Conda for Python Environment Management

    Conda is an open-source package and environment management system used primarily for Python. It allows developers to create isolated environments, manage complex package dependencies, switch between project-specific environments, and resolve version conflicts across Windows, macOS, and Linux.

  9. Overview of Alicia-DL Force Feedback Leader Arm

    The Alicia-DL Force Feedback Leader Arm is a 6-degree-of-freedom (6-DOF) device designed for manual teaching and teleoperation. When paired with the Alicia-M manipulator, it uses teleoperation coordination algorithms to map the leader arm's movements to the manipulator. It supports both position mode and current mode, allowing the operator to perceive force changes experienced by the Alicia-M manipulator through joint torque feedback.

  10. Overview of Synria Desk robotic arm platform

    Synria Desk is a desktop-oriented robotic arm control and collaboration platform. It provides a comprehensive suite of tools for device onboarding, single-arm and dual-arm control, remote teleoperation, custom URDF visualization, firmware management, and account/billing administration.

  11. Overview of Gloria-D Force Control Gripper

    The Gloria-D Force Control Gripper is an advanced version of the Gloria-D series parallel two-finger servo gripper. It builds upon high-precision position control by integrating real-time torque sensing and active force control capabilities. This allows for compliant grasping and force-feedback closed-loop control, making it suitable for precision assembly, handling flexible objects, contact force research, and safe human-robot interaction.

  12. Overview of the Physical AI Infrastructure workflow

    The Physical AI Infrastructure provides a complete end-to-end pipeline for enabling robots to learn and execute tasks from demonstrations. The workflow consists of four primary stages:

    1. Data Collection: Managing collection tasks, configuring devices, and recording real-time robot operations to create training data segments.
    2. Dataset Management: Organizing collected data, previewing segments, assessing quality, performing cloud conversion, and publishing datasets to Hugging Face.
    3. Model Training: Selecting converted data artifacts, configuring the model and training specifications, and tracking results.
    4. Inference Deployment: Loading a trained model, configuring cameras and robotic arms, and executing online inference tasks.