How to Communicate Robot Motion Intent: A Scoping Review

Document Title

How to Communicate Robot Motion Intent: A Scoping Review

Document Information

• Subject Area: Human-Computer Interaction (HCI), focusing on human-robot collaboration and the communication of robot motion intent
• Keywords: intent, motion, robot, collaborative robots, drones, survey, human-computer interaction

Research Background and Problem Statement

• Identified Problems or Challenges:

  • As robots increasingly integrate into daily life and collaborative scenarios, humans need to understand robot motion intent to avoid task failure and improve collaboration efficiency. However, there is currently no unified model to systematize methods for communicating motion intent.
  • There is no consensus on the definition and associated factors of "robot motion intent," and existing terminology and research methods suffer from conceptual ambiguity.
  • Various research approaches, such as using augmented reality (AR) or visualized paths, have not provided a unified classification standard.

• Why This Problem Is Important:

  • Understanding robot intent is crucial for human-robot collaboration and safety, especially in scenarios where robots and humans frequently share physical spaces.
  • Clarifying how robot motion intent is expressed is key to developing explainable robotics and seamless human-robot interaction.

• Research Motivation and Related Work:

  • Traditional robot navigation research has focused more on how robots understand human behavior, but there has been less research on how robots can convey their intent to humans, enabling humans to understand their action plans and goals.
  • Current research emphasizes the integration of technologies such as robot swarms and emotion and intention communication.

Proposed Solution

• Proposed Method or Solution:

  • Through a review of existing literature, an "Intent Communication Model" is proposed to unify and classify methods for expressing motion intent.
  • The model includes three main entities (robot, intent, human role) and multiple dimensions (e.g., intent type, information attributes, and communication location).

• Innovative Aspects:

  • The model is not limited to motion intent but also includes supportive communication types related to motion intent (e.g., attention, state, and instruction).
  • It provides a systematic perspective, organizing the research patterns and practices in the field of robot intent communication.

• Implementation Steps:

  1. Literature Search and Screening: Using databases such as ACM, IEEE Xplore, and ScienceDirect, a systematic keyword search was conducted, resulting in 77 qualifying studies.
  2. Development of the Intent Communication Model: Key elements of intent communication were systematically extracted from the literature and categorized into five main aspects: goal, robot type, intent type, intent information, and human role.
  3. Model Refinement:
     • Intent information was further divided into spatial information and temporal information.
     • Communication locations were categorized, such as on-robot, on-world (environment), and on-human (human accessories).

Research Outcomes

• Specific Outcomes:

  • The Intent Communication Model was proposed and clarified, revealing the classification and relationships of different intents through entity and dimension analysis (robot type, intent type, human role, etc.).
  • Intent communication was categorized in detail (motion intent, attention intent, state intent, and instruction intent), clarifying the functions and interrelations of these types.
  • A systematic perspective was provided on intent information (including spatial and temporal attributes) and its communication locations, summarizing and comparing related experimental studies.

• Advantages Compared to Existing Solutions:

  • Conceptual Unification: Addresses the current issues of unclear terminology and fragmented research directions, providing a classification framework for future studies.
  • High Generalizability: The model is applicable to various robot types (robotic arms, mobile robots, humanoid robots) and different interaction scenarios.

• Experimental or Evaluation Results:

  • For example, experiments using augmented reality (AR) tools to display robot intent significantly enhanced users' trust and sense of safety while reducing task interference.
  • Comparisons of direct projection (on-world), head-mounted devices (on-human), and robot-based displays (on-robot) showed that multimodal approaches improve interaction efficiency.

• Limitations and Future Directions:

  • Limitations:
    • Current research primarily focuses on single robot and single human interaction scenarios, with limited exploration of multi-user or multi-robot collaboration contexts.
    • There is room for further expansion in the classification and attributes of "intent" types.
  • Future Directions:
    • Explore the design, best practices, and technical challenges of intent communication in multi-user scenarios.
    • Establish broader interdisciplinary connections, such as integrating with research on external human-machine interfaces (eHMIs) for autonomous vehicles.
    • Systematically compare the advantages and applicable scenarios of spatially registered information and abstract information.

The above content provides a well-structured and model-based review, laying a theoretical foundation and practical guidance for current and future research on robot intent communication.