1. Small in size, no need for an external control cabinet, easy to integrate

2. Lightweight, weighing only 3.5kg, suitable for various small mobile platforms (quadrupeds, wheeled robots, drones, etc.).

3. AI+: Compatible with cutting-edge embodied intelligence technologies such as ALOHA, Mobile ALOHA, RT-X, VoxPoser, etc. Utilizes large models to understand natural language commands, supports one-click scene scanning, and improves the efficiency of algorithm development with high-fidelity simulation environment.

The power supply voltage range for the AIRBOT Play is DC 24V, with a maximum power of 240W. It is recommended that the power supply device be able to provide an instantaneous maximum current of 10A to support the robotic arm's use in various operating conditions.You can use the power adapter provided in the accessories to connect to the mains power supply, or you can power it through a battery or portable power source using the provided power cable.The power adapter plug is designed to prevent incorrect insertion. Please refer to the following diagram for the correct insertion method:

If it is a CAN interface end effector, you can use the interface at joint 6, which outputs a voltage of 24V and a rated output current of 1.15A. This interface can be used to directly control the AIRBOT Gripper, our self-developed gripper.If it is an RS485 interface end effector, you can use the RS485 interface on the end board, which outputs a voltage of 5V/7V/24V (select output voltage through IAP firmware upgrade), with a rated output current of 1.15A.

The Type-C interface on the end board of the robotic arm can directly connect to cameras.Cameras that have been tested and proven to support internal wiring include INUCHIP, etc. . Additionally, adapter boards for installing cameras at the end of the robotic arm can be provided.

The current version of the AIRBOT Play robotic arm relies on control from a host computer, which can be a PC or a mini PC.

Currently supported simulation platforms include Isaac Sim, Gazebo and MuJoCo.

The grasping algorithms for the gripper include modular grasping algorithms such as GraspNet, CatGrasp, as well as end-to-end planning grasping algorithms like VoxPoser. DEMO and SDK can be provided according to your requirements.

Multimodal perception includes end effector vision, end effector touch, environmental sound, scene SOP instructions, etc. Multimodal perception integrates various environmental information for long-range task execution. DEMOs and SDKs can be provided according to your requirements.