If you feel that you've received this message in error, please click here for more information.

Price: Complete Form Below for a Quote Pre-Order

    Specifications

    Hardware Features

    Move beyond research lab tabletops and into real-world robot training with Mobile ALOHA. It's not just a mobile bimanual robot, it's a powerful platform for collecting and replaying high-quality imitation learning data. Mobile ALOHA is more than just a robot, it's a key that unlocks the potential of imitation learning for real-world bimanual manipulation. Take your research or development to the next level. Contact us today to learn more!

    Zipeng Fu *    Tony Z. Zhao *    Chelsea Finn *   

    Effortless Data Collection:

    Whole-body Teleoperation: Control both the mobile base and ALOHA's bimanual arms intuitively, like puppeteering, to easily record demonstrations.

    Multiple Hours of Continuous Usage: Train for extended periods, tackling complex tasks like cooking a meal or cleaning an entire room.

    Low Cost, Accessible Hardware: Mobile ALOHA leverages the proven ALOHA system with an inexpensive mobile base, making it an affordable and practical choice for researchers.

    Boost Learning with Co-training:

    Leverage Existing Datasets: Bridge the gap between tabletop and mobile manipulation with co-training. Utilize readily available static ALOHA data to jumpstart learning and improve performance on mobile tasks.

    Data Efficiency: Achieve impressive results with minimal data. Train Mobile ALOHA with only 50 demonstrations per task, significantly reducing your data collection effort.

    Fine-tune for Real-World Scenarios: Co-training provides a strong foundation for adapting to diverse environments and challenges, making Mobile ALOHA truly real-world ready.

    Unleash the Potential of Bimanual Mobile Manipulation:

    Master Everyday Tasks: Navigate cluttered spaces, open cabinets, manipulate delicate objects, and even cook a simple meal.

    Go Beyond the Lab: Train robots for real-world applications like home automation, office assistance, and industrial tasks.

    Accelerate Research and Development: Mobile ALOHA provides a platform for exploring the frontiers of bimanual mobile manipulation, opening doors to exciting new possibilities.

    Software

    ROS - Robot Operating System (ROS or ros) is an open-source robotics middleware suite. Although ROS is not an operating system but a collection of software frameworks for robot software development, it provides services designed for a heterogeneous computer cluster such as hardware abstraction, low-level device control, implementation of commonly used functionality, message-passing between processes, and package management.

    CAN - A Controller Area Network (CAN bus) is a robust vehicle bus standard designed to allow microcontrollers and devices to communicate with each other's applications without a host computer. It is a message-based protocol, designed originally for multiplex electrical wiring within automobiles to save on copper, but it can also be used in many other contexts. SCOUT has an internal controller that runs on a CAN-based communication protocol. Users can communicate with the controller over CAN through AgileX’s open CAN bus protocol.

    Hardware


    IP Upgrade - An optional upgrade to BUNKER’s IP rating from the standard 52 to the significantly upgraded 54, protecting it from splashes of water in every direction.


    Christie Suspension Equipped with Multiple sets of Shock absorbers - BUNKER’s Christie suspension, equipped with shock absorbers, keeps the platform stable as it drives over rough terrain at high speeds, giving it the ability to traverse large obstacles with great handling.


    Rails - The top of the robot provides a standardized mounting solution with aluminum T-slot rails, convenient for securely mounting external equipment and sensors or one of the available preconfigured R&D or navigation kits.

    Packages


    Battery Upgrade - (60Ah Expansion Battery) By default, the BUNKER comes with a 48V30Ah battery which allows for a maximum travel distance (without load) of 10km. With the optional 60Ah upgrade, the maximum travel distance increases significantly.

    R&D Kit - The AgileX R&D Kit is a fully integrated solution for robotics research and development. It is equipped with a full suite of sensors to support indoor SLAM, navigation, and vision-based applications. This basic package provides an Nvidia Jetson Nano computer pre-installed with Linux Ubuntu 18.04 & ROS Melodic. A WiFi module, Intel RealSense depth camera, EAI G4 LiDAR, & an LCD display.

    R&D Kit Pro - Like the standard R&D Kit, the AgileX R&D Pro Kit is a fully integrated solution for robotics research and development. It is equipped with a more powerful suite of sensors to support indoor SLAM, navigation, & vision-based applications than its standard version. The R&D Pro package includes the high-performance Nvidia Xavier computer pre-installed with Linux Ubuntu 18.04 and ROS Melodic, a WiFi module, an Intel RealSense depth camera, a Velodyne Puck LiDAR, & an LCD display.

    Autopilot Kit - The AgileX Autopilot Kit is a hardware & software solution that enables autonomous navigation, path planning, & obstacle avoidance without the need for preloaded maps. The Autopilot Kit comes with the following sensors, devices, & computers: Pixhawk 4 controller, high precision LiDAR, RTK GPS antenna, binocular camera, depth camera, a high-performance computer, & WiFi module.

    AutoKit - The AgileX AutoKit is a full-stack and cost-effective autonomous driving development & education kit based on Autoware open-source software. It provides a powerful autonomous driving sensor system that seamlessly integrates into your robot platform. With this software & hardware platform stack, along with comprehensive user guides, the AutoKit empowers educational experts and industry developers to quickly and easily deploy autonomous robots and develop autonomous driving research in various industries. The AutoWare Kit comes with the following sensors, devices, & computers: A computer with 8 cores and 32GB of RAM, Robotsense RO-LiDAR-16 16 channel LiDAR, LCD display, & a USB-to-CAN module.


    Drawings

    Documentation