Oluwatoni Ogunmade

Background

Imọ means knowledge in my mother tongue. This project was initially start to scratch an itch I had on building a maze solving robot. The scope has changed a few times and currently it a robot platform to learn indoor exploration.

---

Design

The robot sports 5 ultrasonic sensors, a wheel speed encoder, a laser distance sensor and an inertial measurement uint. Onboard a raspberry pi and 5 atmel microcontrollers handle the sensor processing and vehicle actuation.

The ultrasonic sensors are solely used for obstacle avoidance while the laser distance sensor is used to map the environment. The IMU and encoder are fused together using an Extended Kalman Filter to provide the SLAM system with better odometry than what can be gotten from each sensor. Currently the robot is capable of passively mapping it's surroundings, which it represents using an occupancy grid. Work is ongoing towards full active indoor exploration with the frontier exploration algorithm being used with the google cartographer library as a possible SLAM algorithm. Follow my progress here on Github where I keep a running list of what needs to be done.