Human Robot Interactions Relying on Human Perception and Intention Awareness
Abstract
Human-robot interaction (HRI) is a rapidly growing field that studies the relationship between humans and robots. A robot's ability to understand and adapt to human intentions and perceptions is key. This study focuses on HRI in the context of mobility assistance, specifically the Smart Walker (SW). The SW can be used as a personal assistive as well as a rehabilitative device. A prototype of such a walker was built equipped with an Inertial Measurement Unit (IMU) with machine learning capabilities, Time of Flight (ToF) range sensors, and pressure sensors on the handles. The sensors measure and interpret the states of the walker and the human user. The walker also includes hydraulic disk brakes on the wheels to respond to control commands. We developed an algorithm with five key operational states that relate human movements to desired actions for the walker. The algorithm utilizes a decision tree framework for classifying various states. A series of experiments were conducted to investigate how the walker could interpret human movement patterns and adjust its function to offer regulated assistance. The SW device opens up the possibility of integrating robots into the human environment as personal companions, allowing people with mobility issues to live independently. The insights gained from this project can be applied to a broader spectrum of HRI applications.
Keywords: Human-Robot Interaction, Mobility assistance, Smart Walker, Machine Learning