PAL Robotics Kangaroo

Name: PAL Robotics Kangaroo
Brand: PAL Robotics

height: 1.58 m · weight: 50–65 kg

Data & sources

Key specifications

Specifications

Dimensions specifications
Height	1.58 m
Weight	50–65 kg

Actuation specifications
Total degrees of freedom	14–40 DoF (platform configuration)
Degrees of freedom per arm	4, 5 or 7 DoF (platform configuration)
Torso dof	2 DoF (yaw and roll)
Degrees of freedom per leg	6 DoF
Actuator control modes	Current; speed; position; impedance; force/torque (optional) PIDs; current; speed; position; impedance; force/torque
Actuators and transmissions	Custom linear actuators with integrated force sensors provide a wide range of motion and high speed and torques at the joints
Arm configuration	4 DoF arms (platform configuration) 5 DoF arms (platform configuration) 7 DoF arms (platform configuration)
Linear actuators	Custom linear actuators with integrated force sensors
Motor placement	Motors close to the body
Leg length decoupling	Decoupling leg length from hip swing motions
Closed loop force torque control	Closed loop force/torque controllers at joint level
Closed loop joint force torque control	Closed-loop force/torque joint controllers

Manipulation specifications
Arm tool flange	ISO 9409-1 interface with quick tool changer ISO 9409-1
Quick tool changer	Interface with quick tool changer
General payload	7 kg (4 DoF arms)
Box picking policy	Box picking
Arm payload 4 extended	7 kg
Arm payload 4 bimanual	30 kg
Arm payload 5 extended	5 kg
Arm payload 5 bimanual	28 kg
Arm payload 7 extended	3 kg
Arm payload 7 bimanual	25 kg
Parallel gripper	Parallel gripper
Parallel gripper configuration	2 fingers, 1 DoF
Seed robotics rh8d end effector	Seed Robotics RH8D
Seed hand configuration	5 fingers, 7 DoF
Inspire robots hand	Inspire Robots RH56F1 – RH56E2
Inspire hand configuration	5 fingers, 6 DoF

Mobility specifications
Whole body inverse dynamics controller	Whole body inverse dynamics controller
Walking policy	Walking
Push recovery policy	Push recovery
Running policy	Running
Jumping policy	Jumping
Stair climbing policy	Stair climbing
Dance policy	Dance
Low moving inertia	Low moving inertia

Compute specifications
Control loop frequency	EtherCAT 2 kHz control loop
Os and middleware	Ubuntu LTS + ROS 2 LTS + PAL OS
Control framework	ros2_control
Robot models	URDF and MJCF model
Simulation and rl tools	MuJoCo and mjlab
Ai development pipelines	Open-source reinforcement learning; motion retargeting; imitation learning
Rl policies library	Walking; push recovery; running; jumping; stair climbing; dance; box picking
Visualisation and diagnostics	WebGUI – web interface; hardware statistics history; Rviz2
Research access level	Access to real-time loop; custom software deployment; ROS 2 API C++/Python
Kinematic whole body controller	Hierarchical IK solver; teleoperation demo with VR kit
Gpu	NVIDIA Jetson GPU (optional)
Ai controller framework	Online control-policy switching
Open source reinforcement learning	Open-source reinforcement learning
Motion retargeting	Motion retargeting
Imitation learning	Imitation learning
Web interface	Web interface
Motor control electronics	In-house electronics for motor control and sensor acquisition
Decentralized electronics	Embedding custom and decentralization

Sensors specifications
Imu update rate	1 kHz update rate; accuracy 0.25° static, 0.5° dynamic
Imu accuracy static	0.25° (static)
Imu accuracy dynamic	0.5° (dynamic)
Vision system	4× RGB-D cameras
Imu ahrs update rate	1 kHz update rate
Wrist force torque sensor	Wrist 6 axis force/torque sensor
Foot force torque sensor	Feet 6 axis force/torque sensor
Leg actuator force sensor	Integrated force sensor
Seed hand tactile pressure sensors	Tactile pressure sensors