AI-based approaches for improving autonomous mobile robot localization in indoor environments: A comprehensive review

Abstract

The adoption of indoor autonomous mobile robot (AMR) has surged significantly, driven by their ability to integrate diverse sensors, maintain low operating costs, facilitate easy deployment, and exhibit superior mobility. Nonetheless, navigating complex indoor environments presents substantial challenges that can impede AMR performance and diminish overall system efficiency. To overcome these obstacles, researchers have concentrated on developing autonomous localization techniques that empower AMR to navigate and execute tasks effectively within intricate settings. Recent advancements in artificial intelligence (AI) applications have profoundly influenced this field, enhancing the control and decision-making capabilities of AMR. This paper offers a comprehensive review of AI-based strategies aimed at improving localization of indoor AMR, including aerial vehicles. We systematically categorize and critically analyze existing research on Simultaneous Localization and Mapping (SLAM)-based methods, odometry-based approaches, and multi-sensor fusion techniques, elucidating the principles and implementations of various AI methodologies. Additionally, we discuss real-time performance challenges associated with AI-based approaches and delineate the distinctions between AI-enhanced localization methods and traditional localization techniques, highlighting the necessity and advantages of adopting AI-based solutions. By clarifying these methodologies, our goal is to enhance their comprehension and promote their widespread adoption within the field. Finally, we discuss emerging research directions and ongoing challenges, providing guidance for future investigations and advancements in the domain of indoor AMR.