Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Advancing Frontiers in SLAM: A Survey of Symbolic Representation and Human-Machine Teaming in Environmental Mapping

Brandon Curtis Colelough

TL;DR

This survey addresses the challenge of robust environmental mapping in SLAM by championing symbolic representations and ontologies within multi-agent systems and human-machine teaming. It synthesizes literature through a PRISMA-guided systematic review, covering semantic SLAM, symbolic reasoning, ontologies, and dynamics across simulation and robotics, and introduces architectural considerations (centralized vs. edge) and the role of control agents. Key contributions include mapping the landscape of symbolic and sub-symbolic SLAM interactions, outlining architectures like SYMBO-SLAM and Onto4MAT, and detailing progress in place recognition, map matching, and dynamic-environment handling. The work highlights the practical significance of transparent, context-aware SLAM for applications ranging from military operations to autonomous navigation, and provides a roadmap for future research in richer ontologies, explainable AI, and hybrid symbolic/sub-symbolic frameworks.

Abstract

This survey paper presents a comprehensive overview of the latest advancements in the field of Simultaneous Localization and Mapping (SLAM) with a focus on the integration of symbolic representation of environment features. The paper synthesizes research trends in multi-agent systems (MAS) and human-machine teaming, highlighting their applications in both symbolic and sub-symbolic SLAM tasks. The survey emphasizes the evolution and significance of ontological designs and symbolic reasoning in creating sophisticated 2D and 3D maps of various environments. Central to this review is the exploration of different architectural approaches in SLAM, with a particular interest in the functionalities and applications of edge and control agent architectures in MAS settings. This study acknowledges the growing demand for enhanced human-machine collaboration in mapping tasks and examines how these collaborative efforts improve the accuracy and efficiency of environmental mapping

Advancing Frontiers in SLAM: A Survey of Symbolic Representation and Human-Machine Teaming in Environmental Mapping

TL;DR

This survey addresses the challenge of robust environmental mapping in SLAM by championing symbolic representations and ontologies within multi-agent systems and human-machine teaming. It synthesizes literature through a PRISMA-guided systematic review, covering semantic SLAM, symbolic reasoning, ontologies, and dynamics across simulation and robotics, and introduces architectural considerations (centralized vs. edge) and the role of control agents. Key contributions include mapping the landscape of symbolic and sub-symbolic SLAM interactions, outlining architectures like SYMBO-SLAM and Onto4MAT, and detailing progress in place recognition, map matching, and dynamic-environment handling. The work highlights the practical significance of transparent, context-aware SLAM for applications ranging from military operations to autonomous navigation, and provides a roadmap for future research in richer ontologies, explainable AI, and hybrid symbolic/sub-symbolic frameworks.

Abstract

This survey paper presents a comprehensive overview of the latest advancements in the field of Simultaneous Localization and Mapping (SLAM) with a focus on the integration of symbolic representation of environment features. The paper synthesizes research trends in multi-agent systems (MAS) and human-machine teaming, highlighting their applications in both symbolic and sub-symbolic SLAM tasks. The survey emphasizes the evolution and significance of ontological designs and symbolic reasoning in creating sophisticated 2D and 3D maps of various environments. Central to this review is the exploration of different architectural approaches in SLAM, with a particular interest in the functionalities and applications of edge and control agent architectures in MAS settings. This study acknowledges the growing demand for enhanced human-machine collaboration in mapping tasks and examines how these collaborative efforts improve the accuracy and efficiency of environmental mapping
Paper Structure (15 sections, 5 equations, 2 figures, 1 table)

This paper contains 15 sections, 5 equations, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Search Process
  4. Inclusion / Exclusion Criteria and Quality Assessment
  5. Paper Selection
  6. Agent search and path planning
  7. Feature extraction
  8. Place Recognition
  9. Multi-Agent System
  10. Map Matching
  11. Ontology and Symbolic Reasoning
  12. Sub-symbolic and Symbolic SLAM in Dynamic Environments
  13. Simulated Environment
  14. Robotics
  15. Conclusion

Figures (2)

  • Figure 1: A literature review of existing architectures surrounding the concept of the proposed SYMBO-SLAM architecture was conducted. This literature review focused using symbolic reasoning through ontological design to create contextual maps of an environment. The Venn diagram in the centre links the majority of concepts required for the implementation of the proposed system in simulation and on hardware. The box below lists the papers directly applicable to the use of symbolic reasoning in the SLAM domain
  • Figure 1: The 6 search terms were queried through the 9 databases as described above. The number of pieces of literature returned from each query is shown in the table above. Note also that results that returned above 30 literature pieces were further filtered by year published for reduction