design teleoperation systems

To address trajectory planning failures of autonomous vehicles in complex scenarios—leading to restricted operational design domains (ODDs)—this paper proposes a lightweight, loosely coupled remote assistance architecture based on arbitration graphs. Without modifying existing system code, the approach introduces a remote human intervention layer enabling real-time planning-level overrides and dynamic ODD expansion. A modular arbitration graph framework coordinates human–machine decision-making, balancing operator workload and system safety. Evaluation on two representative use cases in simulation demonstrates the architecture’s efficacy in supporting remote constraint adjustment and out-of-ODD trajectory generation, while ensuring compatibility, real-time responsiveness, and engineering deployability. The core contribution is the first “plug-and-play planning-layer” remote collaboration paradigm, establishing a novel pathway for dynamic ODD evolution.

Novice operators face significant challenges in teleoperating robots, including high operational complexity, poor safety guarantees, and limited cross-platform compatibility—factors that hinder both robot learning and efficient data collection. To address these issues, we propose a low-cost (<$1,000) teleoperation system featuring a novel real-time virtual arm visualization mechanism that enables pre-execution rehearsal of commands, supporting seamless toggling between action preview and execution. The system leverages lightweight motion mapping, real-time virtual rendering, and low-latency interactive feedback, requiring no specialized hardware and ensuring compatibility with mainstream robotic arm platforms. Evaluated on five dexterous manipulation tasks, our approach outperforms existing methods in task success rate and operator efficiency. It substantially reduces the learning curve for novice users and improves operational safety. All source code and deployment documentation are publicly released under an open-source license.

Autonomous vehicles lack robust remote takeover support systems for safe operation on public roads. Method: This study proposes a remote takeover control center framework designed for real-road validation. Through task analysis and role-function mapping, it rigorously defines responsibility boundaries between remote operators (focused exclusively on driving接管) and fleet managers (responsible for scheduling and anomaly response). It introduces a novel, standardized state diagram covering all takeover scenarios, explicitly modeling vehicle-operator collaborative state transitions to ensure regulatory compliance and adherence to human-machine interaction prohibitions. Contribution/Results: The resulting state-machine-driven workflow and deployable architecture have enabled compliant testing and validation of multiple vehicle platforms on open roads. This work establishes a reusable, verifiable technical foundation for vehicle–road–cloud integrated remote assistance systems.

Remote intervention for autonomous vehicles in complex scenarios—such as road construction—poses significant human factors challenges, necessitating effective human–machine interface (HMI) designs that balance usability, cognitive load, and operational reliability. Method: This study conducts the first human-centered comparative experiment on remote control interaction paradigms, systematically evaluating three HMI approaches—path planning, trajectory guidance, and waypoint guidance—using a within-subject design on a real-time multi-vehicle simulation platform. Performance was assessed via the System Usability Scale (SUS), NASA-TLX cognitive workload, task completion rate, and subjective satisfaction. Contribution/Results: Path planning significantly outperformed the other two paradigms (SUS = 78.2; highest task completion rate), while trajectory guidance yielded the lowest success rate. All 23 participants unanimously preferred path planning. The findings reveal critical trade-offs in human–autonomy collaboration under complex roadwork conditions, providing empirical evidence and design guidelines for high-reliability remote monitoring HMIs.

To address key challenges in human–machine interfaces (HMIs) for underwater remotely operated vehicles—including low immersion, unintuitive control, lack of cross-platform standards, and insufficient shared autonomy—this study systematically reviews over 100 state-of-the-art approaches, identifying turbid-water perception degradation and communication latency as fundamental bottlenecks. We propose a novel multidimensional HMI evaluation framework integrating human-centered metrics (e.g., intuitiveness, situational awareness) with robotic performance criteria (safety, accuracy, efficiency). Furthermore, we introduce a bidirectional intelligent collaboration paradigm supported by multimodal feedback, incorporating real-time gesture recognition, VR/AR-based rendering, large language model–driven natural language interaction, ultra-low-latency communication, and multi-source underwater state estimation. Finally, we distill six open research questions and outline an interdisciplinary roadmap, providing theoretical foundations and design guidelines for next-generation intelligent deep-sea intervention systems.

This study addresses the challenges of high cognitive load on single operators and substantial coordination overhead in multi-operator setups for multi-arm teleoperation tasks. The authors propose a human–robot collaborative teleoperation framework wherein a human directly controls two master arms, while two auxiliary arms are autonomously managed by a training-free multimodal large language model (MLLM) agent to execute subtasks, with real-time intervention enabled via voice commands. This approach pioneers the integration of MLLM agents into multi-arm teleoperation for data collection, achieving decoupled control spaces and natural human–robot interaction. The system maintains high operational efficiency while significantly enhancing scalability. Experimental results demonstrate that the proposed method achieves data collection success rates and efficiency comparable to those of expert two-human teams, and the collected data effectively supports downstream training of multi-arm collaborative policies.

为应对多样化生产带来的挑战，本文提出一种基于CAD模型的编程和执行系统，通过动态参数化的行为树控制结构实现人-机器人-起重机协同任务。

This study addresses the lack of a unified requirements framework in current human-AI collaboration tasks, which hinders the design and evaluation of complex cooperative systems. By systematically reviewing academic literature, industry standards, and regulatory guidelines, this work proposes the first comprehensive requirements taxonomy encompassing six high-level categories and twenty-one subcategories. The framework integrates fragmented knowledge and clarifies critical dimensions such as information provision, relationship control, and decision support. Through an iterative process of requirement extraction, classification, and expert validation, the authors derived 361 requirements from 14 sources to construct the taxonomy and further demonstrated its applicability on an independent corpus containing 448 requirements. The resulting framework received endorsement from five domain experts.

GHOST系统通过实时整合机器人摄像头数据和基于学习的场景补全，解决了单人同时操控两台移动操作机器人的问题，提高了操作成功率和效率。

This study addresses the challenges of hazardous construction environments and the limited autonomy of humanoid robots by proposing a hybrid teleoperation platform based on the Unitree G1. We introduce a novel architecture integrating extended reality (XR) with foot-pedal controls, enabling a single operator to simultaneously manipulate upper-body tasks and lower-body navigation. This approach effectively resolves complex whole-body coordination challenges while ensuring high-quality data acquisition. Experimental evaluations demonstrate that the system achieves success rates of 100% and 80% in tool transportation and surface painting tasks, respectively. Although execution efficiency remains below human performance baselines, these results substantiate the platform's significant potential for deployment in dangerous construction scenarios.