servolab

Control Feetech serial bus servos (STS3215 and friends) from Python. Made for teaching robotics: readable code, SI units, a simulator and a web app to look at your motors. Developed for the robotics seminar of the TUM Chair of Microtechnology and Medical Device Technology (MiMed).

import servolab

bus = servolab.connect()        # finds the Waveshare USB adapter and all motors
m = bus[1]                      # the motor with ID 1 (an ID, not a list index)

m.mode = "position"
m.torque_enabled = True
m.position = 0.5                # rad
print(m.position, m.current, m.temperature)

No hardware? servolab.connect("sim") gives you three simulated motors.

Installation

pip install "servolab[all]"

or, in a uv project:

uv add "servolab[all]"

Python 3.10 or newer, on Windows, macOS and Linux. The core (motors, IMU, simulated motors, network) only needs numpy and pyserial; everything else is optional:

Extra	Adds	Packages
scope	the ServoScope web app	nicegui
plot	rec.plot(), LivePlot	matplotlib, rerun-sdk
kinematics	the robotics toolbox (used in courses)	roboticstoolbox-python
sim	simulated robots with physics	mujoco, rerun-sdk
all	everything above

A computer that only shares its motors over the network, e.g. the Raspberry Pi of a mobile robot, needs just pip install servolab.

Where to go next

• Getting started: your first lines of code, moving motors, loops and plots.
• Examples: small programs that go step by step.
• Python API: every function and property, with units.
• ServoScope: look at your motors in the browser.
• Hardware: wiring, IDs, and what was checked on real motors.