Sphero RVR

This projects extends your Sphero RVR with different hardware modules:

Buzzer to talk like R2D2
Ultrasonic modules
PiCamera

Includes different software addons:

Create sound using buzzer
Measure distances using the ultrasonic sensors
Security layer that is ALWAYS active to avoid crashes and cliffs
Object detection using asyncio with TFLite 2.1

Sample applications included in this repo:

Follow me
(Random?) Autonomous driving
Create maps of rooms
Drive to a room after a map is created and take pictures
Security robot -> check if there is any anomaly in any of the rooms (unknown faces etc.)

Note: This project is WIP and thereofe we can not avoid bugs etc. Also the security layer can avoid crashes only with some probability, because the underlying hardware and software that we use is not a hard realtime system.

Setup

Software

Clone this repository and execute ./install.sh Note that we assume for the installation that the cloned folder name is sphero-rvr. Note also that in the directory above the sphero python sdk is installed.

To add new packages run pipenv install PACKAGE

Hardware

For the hardware we need to connect the raspberry pi 4b to the sphero rvr, install the raspberry camera, install 3x ultrasonic sensors and the buzzer. To test whether your hardware works corretly, you can use the test scripts of test/*.py

Ultrasonic

The RPI gpio's work with 3.3V, the ultrasonic sensors with 5V. Therefore the echo of the ultrasonic is downscaled to 3.3V via two resistors R1 and R2: R_2 = 3.9k and R_1 = 2.2k and therefore as V_out / V_in = R_2 / (R_1 + R_2)

such that V_in ~ 3.2V:

              +
              |
              |
        Echo  +--------+
              |        |
              |      +---+
              |      |R_1|
              |      |   |
 Ultrasonic   |      +---+
   Sensor     |        |
              |        +----------+  GPIO_IN Raspberry Pi
              |        |
              |      +---+
              |      |R_2|
              |      |   |
              |      +---+
              +        |
                     +---+

Sphero RVR connection

Ultrasonic sensors

Speakers

Camera

Description

Security Layer

Every example can access functionality of the robot only through the security layer. This layer ensures that no crash etc. can happen. Therefore autonomous driving etc. can focus on this funcitonality and not on security relevant stuff.

Tipps for devs

To develop on the Sphero RVR I use VsCode (on a dev machine) directly debug the raspberry on the robot over ssh.

References

[1] Ultrasonic RVR example - https://sdk.sphero.com/docs/samples_content/raspberry_pi/python/ultrasonic_rvr_sample/
[2] Install Sphero RVR SDK - https://sdk.sphero.com/docs/getting_started/raspberry_pi/raspberry_pi_setup/
[3] Raspberry GPIO pins - https://www.raspberrypi.org/documentation/usage/gpio/
[4] Tool to create laser-cut box - https://www.makercase.com/#/kerfbox

peerdavid/sphero-rvr