Maze Robot
Maze Robot
Embedded Files
This project was a part of a Mechatronics Principles course in which we designed and built maze traversing robots from scratch. The robot needed to be capable of obstacle avoidance, localization, and object pick-up/drop-off.
This project was a part of a Mechatronics Principles course in which we designed and built maze traversing robots from scratch. The robot needed to be capable of obstacle avoidance, localization, and object pick-up/drop-off.
Sensors
Sensors
The robot was equipped with two types of sensors, ultrasonic (USS) and time of flight (ToF). Our team rigged the ToF sensor on a turntable on the top of the robot to act as a cheaper homemade lidar, however, in practice we ended up mainly relying on the USS. The USS were mounted in parallel pairs to be able to determine the distance and angle of adjacent walls. The information from these sensors could be used to keep the robot centered in maze corridors and avoid obstacles.
Manufacturing
Manufacturing
The robot was a combination of laser cut and 3D printed parts, featuring a modular design with the ability to insert more "levels" as needed. 3D printing was used for brackets and other mechanical parts. We minimized messy wires by soldering some wires and components.
The robot was a combination of laser cut and 3D printed parts, featuring a modular design with the ability to insert more "levels" as needed. 3D printing was used for brackets and other mechanical parts. We minimized messy wires by soldering some wires and components.
Results
Results
Our design was able to effectively complete all tasks. Using an Arduino microcontroller, the robot was able to effectively process sensor data to avoid obstacles and localize using a landmark localization method. The task that the robot struggled with the most was block pickup, as the USS was not reliable enough to always detect the block and the micro servo sometimes struggled to generate enough torque.
Our design was able to effectively complete all tasks. Using an Arduino microcontroller, the robot was able to effectively process sensor data to avoid obstacles and localize using a landmark localization method. The task that the robot struggled with the most was block pickup, as the USS was not reliable enough to always detect the block and the micro servo sometimes struggled to generate enough torque.
Check out the video below for more project details.
Check out the video below for more project details.
Maze Robot Presentation.mp4
Page updated
Google Sites
Report abuse