ALESSANDRO TIOZZO | MECHATRONIC ENGINEERING STUDENT
BulbaBot, the plants' guardian
This hexapod programmed in C++, Python, ROS and Bash, has as mainly purpose to walk torwards the sunlight in order to take care of the plant which is on its back, dodging possibles obstacles on its way. It uses a telegram bot to communicate with the user and it is possible to control it remotely.
It has been developed in a team for Robotic exam, and now I am trying to develop it by myself, using feedbacks and a front camera, to perform inverse kinematics on its legs and trajectory planning to make its movements more natural.
3D printed parts
The overall hardware of the robot is composed by 3D printed parts. The motors used are 180° servo MG996R, controlled with arduino and Raspberry.
The main control of the robot is managed through information exchange between Arduino and Raspberry using ROS nodes.
First movements
The first walk of the hexapod was extremely rough and slow. Has been discoreved during the testing that some parts needed a new design to improve stability.
The procedure of the walk is evaluated through inverse kinematics and it is calculated one time only at the startup.
Initial results
The first good working prototype was able to detect obstacles and avoid them, turning its head in order to detected an higher range of visibility through its proximity sensor. Then the walk process has been improved.
BulbaBot 1.0
The final product got some new features, the more visible is the plant on the robot's back with a humidity sensor inside the pot and the light sensor on its head to follow the sunlight.
Last but not least the weak up process needed a routine in order to get the robot in walking position.
Remote Control
In the mean time, the remote control through the telegram bot has been implemented. Another ROS node get the information from the bot from Raspberry and send commands directly to Arduino which controls the servos.
New ideas
After the Robotic exam I decided to upgrade all the robot functionality, to change some element's design and develop more robust alghoritms, starting with getting rid of Arduino and building my own shield.
Topology Optimization
During the exam period I used my free time for learning new skills as 3D CADs usage and basic concepts of topology optimization, here are the results of the robot assembly in Inventor after the optimization of the femur.