Indoor autonomous robot navigation based on semantic information and sequential directional commands

• المؤلف الرئيسي: Heng, Kuan Kian
• مؤلفون آخرون: Zhong Zhaowei
• التنسيق: Final Year Project
• اللغة: English
• منشور في: Nanyang Technological University 2021
• الموضوعات: Engineering::Mechanical engineering::Robots
• الوصول للمادة أونلاين: https://hdl.handle.net/10356/150800
• المؤسسة: Nanyang Technological University

This work proposes a Semantic Instructional-based map-less autonomous navigation system in indoor corridor environment based on Junction Detections and a set of Sequential Instructions given by a human. This contribution brings forth the interactions between an agent and a human, to employ on the collaborations to complete a navigation mission. A set of Sequential Instructions is given prior to the navigation task. With the knowledge of junction recognition and each specific executable action at each junction, the agent has the capability to navigate in any corridor environment, from small to large areas. The system architecture comprises of: “Environmental Processing”, “Junction Detection” and “Decision Making and Semantic Path Planning”. Extensive studies have shown some remarkable pointers in our approach such as: generalization in unseen environments, map-less navigation in large area environments and real-time algorithms. Experimental results present the robustness in the performance on a real robot in real-world environments.