robot car ecosystem 🤖🚗
The idea behind robby-five is to build different components in a robotic/autonomous (toy) car ecosystem.
Inspired by johnny-five the main controlling unit will run on higher level languages like nodeJS or python. The underlying hardware components will be controlled by arduino hardware.
The base setup is the elegoo smart robot car kit v1 or v2. However the components on it only allow to build really basic robots. To make it usable for some higher level automation it needs the base needs to be extended with:
- contents
- project ideas
- examples
- [firmata-servo](#firmata-servoexamplesfirmata-servo)
- [j5-chartist-sonar](#j5-chartist-sonarexamplesj5-chartist-sonar)
- [raspi-led](#raspi-ledexamplesraspi-led)
- [raspi-sonar-servo](#raspi-sonar-servoexamplesraspi-sonar-servo)
- troubleshooting
a list of ideas how the standard set of the robot can be improved to become a bit smarter.
Hook the car up with at least 5 sonar sensors to allow better areal awareness. 
The idea is to interface the sensors via an additional arduino that exposes the individual sensors over I2C.
re-implement the default smart robot car program using raspi & nodeJS as control device. Use sonar for obstacle detection:
- move straight until obstacle <20cm distance
- move sonar to right/left using the servo
- turn robot in the direction with most distance to next obstacle
- repeat
some issues with the base robot car should be improved when it comes to movement:
DC engines do not allow for proper odometry due to varying speed, maybe use steppers
This has been implemented in projects/firmata-sonar-map wheels on each side are controlled in pairs, using individual wheel control might improve maneuverability e.g. only stopping rear wheel to turn
before:
the original setup comes with one motor driver that allows two seperately controlled outputs. this is set up to drive the wheel pair on either side in sync. 
after:
adding an additional motor driver would allow to drive each wheel individually. this would require 2 additional 4 additional pins on the µc to be used. 
wheels have high grip which makes it hard to turn. having lower friction (especially lateral) would allow easier turns.
turns out that turning is only hard for low wheel speeds, increasing to proper speeds lets the robot turn just fine 👯
although quite expensive (100€ per set) the mecanum wheels would allow amazing 2D movements in any direction incl strafing. control logic would be a bit trick probably as well as possible negative influences on odometry solutions
evaluate using 2 driven wheels in front and one undriven wheel in the rear for stability. this would be a cheap solution for the masses. unsure what negative side effects this would have as opposed to 4 wheels.
Set up a stationary camera that tracks the robot. Locate it on the map and send use this information to make the robot move within the map.
come up with an idea how to use the sonar sensor (possibly rotating) to map a room by driving through it.
first implementation of a radar chart with rotating 180° servo is available in the firmata-sonar-map project
use odometry, inertial sensors, sonar/lidar feedback to allow for a control loop system aware of its past movement.
attach laser diodes & laser sensor to the robot. build a remote control web UI to maneuver the robot. attach LED strips to indicate health bars (or on the web UI). use your laser beam to hit other robots
There is a list of examples on how to use the underlying components within this project
Shows how to use johnny-five or firmata.js to control a servo.
Capture HC-SR04 sonar distance readings via johnny-five and chart them on a website.
Use raspberry pi IO to play with LEDs using johnny-five / raspi-io
raspberry as control server for two attached arduinos
for some reason jessy doesn't ship with git installed which the serialport module needs during its make.
apt-get install gityes it happens.. guess it's something with the compilation use npm install instead