Automatic Detection of Myocontrol Failures Based upon Situational Context Information

Abstract

Myoelectric control systems for assistive devices are still unreliable. The user's input signals can become unstable over time due to e.g. fatigue, electrode displacement, or sweat. Hence, such controllers need to be constantly updated and heavily rely on user feedback. In this paper, we present an automatic failure detection method which learns when plausible predictions become unreliable and model updates are necessary.Our key insight is to enhance the control system with a set of generative models that learn sensible behaviour for a desired task from human demonstration. We illustrate our approach on a grasping scenario in Virtual Reality, in which the user is asked to grasp a bottle on a table. From demonstration our model learns the reach-to-grasp motion from a resting position to two grasps (power grasp and tridigital grasp) and how to predict the most adequate grasp from local context, e.g. tridigital grasp on the bottle cap or around the bottleneck. By measuring the error between new grasp attempts and the model prediction, the system can effectively detect which input commands do not reflect the users intention. We evaluated our model in two cases: i) with both position and rotation information of the wrist pose, and ii) with only rotational information. Our results show that our approach detects statistically highly significant differences in error distributions with p < 0.001 between successful and failed grasp attempts in both cases.

Advancements and Challenges in Robot Grasping and Manipulation for Aspiring Researchers

“…In the realm of human-robot interaction (HRI), a focus on intuitive interfaces and cooperative task execution has spurred the development of collaborative robotic systems that can work alongside humans effectively and safely. By enhancing communication, feedback mechanisms, and shared control interfaces, these novel approaches aim to facilitate seamless collaboration between humans and robots in various domains, including manufacturing, healthcare, and assistive technologies [20], [21].…”

Human-Robot Interaction With Robust Prediction of Movement Intention Surpasses Manual Control

“…The previous study related to intention detection and user assistance typically relies on myoelectric or EEG recordings which have limitations associated with a low signal-to-noise ratio (reviewed in Lobo-Prat et al, 2014 ; Losey et al, 2018 ). Conversely, we estimate the motor intention directly from the user kinematics, which has been shown to be effective in our previous study (Heiwolt et al, 2019 ). Furthermore, LQR controllers have previously been employed for solving the problem of user assistance (Borner et al, 2015 ) and correcting user-given input (Medina et al, 2012 ; Moualeu and Ueda, 2014 ).…”

Aging Touch: Systematic and Unbiased Presentation of Tactile Stimuli

“…The system is composed of i) a three-DOF parallel robot arm (Force Dimension Delta 3), ii) a six-channel force-torque sensor (NI-DAQ), and iii) a stepper motor (Arduino Uno). The first two components are controlled by a robotic software framework, Golem, which provides control and planning of the robotic arm, as well as the processing of the FT sensor signals (see for further details [5,11,10,9,13,3,1,12,8,7,6,2,4,7]).…”