Built for research
Backdrivable by design
Quasi-direct-drive motors on every joint. Push the arm by hand, feel the controller, and learn how real robots react to contact.
Impedance control
Command position, velocity, feedforward torque and per-joint stiffness and damping in a single call, at the bus rate.
Gravity compensation
Computed from the URDF with Pinocchio. Switch it on and the arm floats, ready for teaching and free-drive experiments.
Python & C++
The same API in both languages: ulixpy for fast prototyping, ulixarm_cpp for realtime loops without allocations.
Parallel gripper
90 mm stroke, controlled by width, ratio or grip strength, plus an Intel RealSense D405 depth camera in the box.
Open model
Full URDF with visual and collision meshes and dynamic parameters. Use it in any simulator or planner.
From box to motion in minutes
- Clamp and wire. Mount the arm on a table and connect the Power Box to your PC over USB-C.
- Install with one command. The installer builds the API and sets up a ready-to-use Python environment.
- Run your first program. Enable gravity compensation and guide the arm by hand.
- Python
- C++
import numpy as np
from ulixpy import UlixArm
robot = UlixArm(gripper=True)
robot.connect("/dev/ttyACM0")
robot.enable()
robot.refresh()
zero = np.zeros(6)
try:
while True:
# kp = kd = 0: the arm floats, move it by hand
robot.control(zero, zero, zero, zero, zero, True)
finally:
robot.hold()
robot.disconnect()
#include <ulixarm_cpp/ulixarm.hpp>
int main()
{
ulixarm::UlixArm robot(/*gripper=*/true);
robot.connect("/dev/ttyACM0");
robot.enable();
const auto zero = ulixarm::Joints::Zero();
// kp = kd = 0: the arm floats, move it by hand
while (true)
robot.control(zero, zero, zero, zero, zero, /*gravity=*/true);
}
Ready to start?
Dive into the documentation for the Python and C++ APIs.