Dear Community:
The Openpilot Community needs your help to continue to keep's lights on and support the path to Workbench v0.2. With only a one person crew and without commercial funding it's not always financially feasible to fund these services out of pocket. Please consider becoming a Patreon supporter of the Openpilot Community project and receive exclusive perks and benefits!
Become a patreon Learn more Maybe Later

DISCLAIMER: This website is maintained by openpilot users and contributors. The content found here (including Workbench and Oppey the Bot) are not approved, supported, affiliated or funded by, Inc. Please DO NOT contact, Inc. staff about these things as they WILL NOT provide support. For support on openpilot,, oppey, workbench, etc. please contact a community member on Discord.

Vasily Tarasov
Sep 07, 2018·2 min read

Custom DIY Steering Actuator


Demo video

The brain of the system is a legacy NEO hardware. It's capable of both lane keep assist and adaptive cruise control, but I only used it for LKA. Check out project's source code, which is a fork of now out-of-date v0.2.9 release of openpilot. I added rotary encoder to measure steering wheel angle, and control the motor driver (one pin for direction and one for PWM speed). Extra inputs/outputs to the microcontroller are put into a standalone connector:

Rotary encoder input pins are PB15 (configures as an external interrupt 15) and PB14. Motor driver's direction pin is connected to PC6. PC7 is configured similar to fan speed controller: it outputs PWM with variable duty cycle:

I should've learned how to use EagleCad and have ordered a modified board. As a one-time thing, I just soldered wires directly to the microcontroller. It's an easy way to burn the STM32, by introducing an accidental short between pins. Don't do what I did, just find someone to modify and print proper PCB boards for you instead. You might want to add 10k resistors inline to all GPIO inputs/outputs (instead of having them outside of the board), and put a hardware debouncer for the rotary encoder on the board as well. Note that the diagram for the debouncer says 5v, which is misleading. I actually burned one of the boards by putting 5v to rotary encoder's input pins, exceeding 3.3v limit. Power rotary encoder by 3.3v power line instead.