Safety in human-robot interaction: a survey
This paper reviews safety challenges in physical human-robot interaction and categorizes methods to prevent physical injury.
The authors survey the literature, dividing safety approaches into active (sensor-based control, emergency stop) and passive (mechanical design, soft coverings), and discuss injury assessment metrics.
They show that combining force/torque sensors with adaptive control can significantly reduce risks, and existing standards (e.g., ISO) provide a solid framework.
The paper focuses mainly on physical safety and gives less attention to psychological safety (e.g., fear of robots). It also does not cover deep learning methods for human state detection.
The concepts in this paper are used in designing all industrial and service collaborative robots. In the humanoid service robot project, implementing collision detection and force limiting based on these principles is essential for safe user interaction.
📇 Summary flashcard — 13 analytical fields for this paper
خلاصه
This survey paper reviews safety methods in human-robot interaction, covering active and passive approaches to prevent injury.
نمای سریع
A reference for collaborative robot safety with a systems perspective.
یافتههای کلیدی
They show that combining force/torque sensors with adaptive control can significantly reduce risks, and existing standards (e.g., ISO) provide a solid framework.
هدف
To categorize safety methods and provide a framework for designing safe robots.
روش
The authors survey the literature, dividing safety approaches into active (sensor-based control, emergency stop) and passive (mechanical design, soft coverings), and discuss injury assessment metrics.
نتایج
Introduction of injury metrics and emphasis on combining sensors and adaptive control.
نتیجهگیری
Safety is a multifaceted issue requiring a holistic approach from mechanical design to software control.
مفاهیم کلیدی
safety، collaborative robots، HRI، standards، force control
مطالعهی بیشتر
https://www.sciencedirect.com/science/article/abs/pii/S0921889007001328
تحلیل
This is a foundational paper in robot safety and is widely cited as a reference.
محدودیتها
The paper focuses mainly on physical safety and gives less attention to psychological safety (e.g., fear of robots). It also does not cover deep learning methods for human state detection.
کارهای آینده
The authors stressed the need for new standards for complex interactions and methods to detect intentional contact.
کاربرد عملی
The concepts in this paper are used in designing all industrial and service collaborative robots. In the humanoid service robot project, implementing collision detection and force limiting based on these principles is essential for safe user interaction.
◀ References (this paper draws on) (0)
No references recorded.
Cited by (0) ▶
No paper cites this one yet.
Prerequisite learning path for this paper