Roboter-Autos mit dem Raspberry Pi

Roboter Autos mit dem Raspberry Pi - 2019

The Discoverer is a remote controlled robot-car based on a Raspberry Pi. The Discoverer is the biggest robot-car I build. The robot is equipped with a metal detector, a pan & tilt Raspberry Pi camera and a Raspberry Pi Sense-HAT with a magnetometer and gyroscope. I developed the robot as an early beta model for testing the integration of the data generated in combination of a GPS receiver and magnetometer. I learned a lot during the building phase about robotics in general and how to build a chassis. The wheels I actual use are too small for rough terrain. The following picture shows the robot-car driving around a parking lot of the university next to me. As you see in the picture the wheels are very small and it was hard for the robot to drive on the gravel underground.

I was able to control the robot-car over a distance of 350 meters. It was a lot of fun to drive around and to detect some old screws and cables with the metal detector. Everything went very well and prove of the concept was a success. Hopefully I will find in the near future a treasure to re-finance my robot-car hobby 😉

Discoverer with gps antenna

Discoverer with gps antenna


  • GPS supported way point driving
  • metal detector (The metal detector is form Gary)
  • remote controlled over a WIFI connection
  • build in access point
  • live video streaming
  • web interface

Component Overview

If you are thinking about building your own robot-car I created an overview of the components I typically use in my robots.

The component list is available here: component list

GPS support for autonomous driving

With the NAVILOCK GPS NL-602U USB GPS receiver the robot-car is able to find out his position in the an open field outside the city. With the combination of the Raspberry Pi Sense-HAT the robot is able to determine the compass direction his front is facing too. This enables the robot-car to directly turn in the direction the next GPS waypoint is located without driving around to determine the north direction the robot is facing to. To activate the GPS receiver was very easy because it is directly supported by Linux / Raspbian.

GPS software – to control the self-driving robot-car

I developed the software I used to control my robot-car via GPS by my own. The main gps-drive-robo-car.py program is developed in Python. The main program needs three modules to control the l298N H-Bridge, the led matrix from the Raspberry Pi Sense-HAT and a module to calculate the north direction.

self-driving robot program overview

self-driving robot program overview

With the magnetometer and gyroscope from the Raspberry Pi Sense-HAT the robot-car is able to control the direction the front is facing too and to move the robot car in the direction where the next waypoint is located.


The programs are available for download from my download page: download page

Raspberry Pi Sense-HAT

To configure the Raspberry Pi Sense-HAT was not so easy. After the successfully configuration the magnetometer inside the Sense-HAT is doing a perfect job. I am using two Raspberry Pi Sense-HATs in my robot-cars and I am very happy with the Sense-HAT module.

I developed a Python program for the Discoverer which enables the robot to drive from one GPS waypoints to the next one and to use the metal detector to detect some metal or hopefully a treasure. This is the way I like it to hunt for treasures. At the end a log file is generated with the GPS coordinates where the metal detector detected some metal in the ground.

Raspberry Pi self-driving ca

Raspberry Pi self-driving car

Pan&Tilt camera kit

With the camera pan&tilt kit the Discoverer is able to stream a video with a 360° round view. This make the steering very easy in your command center. With the web interface I programed for the Discoverer I am able to control the Pan&Tilt camera as well the four DC motors.

Discoverer Video

The video shows the first test run with active metal detector. The test run was very bumpy and the robot lost a wheel.

(Visited 7,238 times, 2 visits today)