Beyond Barriers: Comparative Insights into Machine Learning Algorithms for Autonomous Mobile Bots in Indoor Environments

In this paper, the development of an Indoor Autonomous Mobile Bot has been introduced by leveraging an Ultrasonic Sensor and Infrared Sensor, in tandem with a Controller. The Ultrasonic Sensor records the obstacle distance data, while the Infrared Sensor measures the velocities of both wheels. Within this investigative framework, three Machine Learning (ML) algorithms—Adaptive Stochastic Gradient Descent Linear Regression (ASGDLR), Adaptive Coordinate Descent Logistic Regression (ACDLoR), and Adaptive Stochastic Gradient Descent LARS Regression (ASGDLARS)—are implemented for the explicit objective of negotiating obstacles within constrained spatial confines. The findings encapsulate simulation outcomes that scrutinize diverse facets of the Confusion Matrix, alongside the computational derivation of obstacle avoidance percentages in the context of singular obstacle scenarios across varying locomotive speeds.