Robust Multi-User In-Hand Object Recognition in Human-Robot Collaboration Using a Wearable Force-Myography Device

Bamani, Eran; Kahanowich, Nadav D.; Ben-David, Inbar; Sintov, Avishai

doi:10.1109/lra.2021.3118087

Cited by 8 publications

(5 citation statements)

References 75 publications

(82 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Vision can also be integrated in order to provide a complete solutions when a line-of-sight is not continuous. Similarly, one could integrate Force-Myography [51], [52] or EMG [32] to have additional information about the pose of the palm and fingers. An additional system on the other arm could enable the classification of tasks performed by the human towards assistance by a robot.…”

Section: Discussionmentioning

confidence: 99%

Learning Human-Arm Reaching Motion Using a Wearable Device in Human–Robot Collaboration

Kahanowich,

Sintov

2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

Many tasks performed by two humans require mutual interaction between arms such as handing-over tools and objects. In order for a robotic arm to interact with a human in the same way, it must reason about the location of the human arm in real-time. Furthermore, to acquire interaction in a timely manner, the robot must be able predict the final target of the human in order to plan and initiate motion beforehand. In this paper, we explore the use of a low-cost wearable device equipped with two inertial measurement units (IMU) for learning reaching motion in real-time applications of Human-Robot Collaboration (HRC). A wearable device can replace or be complementary to visual perception in cases of bad lighting or occlusions in a cluttered environment. We first train a neural-network model to estimate the current location of the arm. Then, we propose a novel model based on a recurrent neural-network to predict the future target of the human arm during motion in real-time. Early prediction of the target grants the robot with sufficient time to plan and initiate motion during the motion of the human. The accuracies of the models are analyzed concerning the features included in the motion representation. Through experiments and real demonstrations with a robotic arm, we show that sufficient accuracy is achieved for feasible HRC without any visual perception. Once trained, the system can be deployed in various spaces with no additional effort. The models exhibit high accuracy for various initial poses of the human arm. Moreover, the trained models are shown to provide high success rates with additional human participants not included in the training of the model.

show abstract

Section: Discussionmentioning

confidence: 99%

Learning Human-Arm Reaching Motion Using a Wearable Device in Human–Robot Collaboration

Kahanowich,

Sintov

2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…Vision can also be integrated in order to provide a complete solutions when a line-of-sight is not continuous. Similarly, one could integrate Force-Myography [42,43] or EMG [17] to have additional information about the pose of the palm and fingers. An additional system on the other arm could enable the classification of tasks performed by the human towards assistance by a robot.…”

Section: Discussionmentioning

confidence: 99%

Robust Classification of Grasped Objects in Intuitive Human-Robot Collaboration Using a Wearable Force-Myography Device

Kahanowich

Sintov

2021

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Many tasks performed by two humans require mutual interaction between arms such as handing-over tools and objects. In order for a robotic arm to interact with a human in the same way, it must reason about the location of the human arm in real-time. Furthermore and to acquire interaction in a timely manner, the robot must be able predict the final target of the human in order to plan and initiate motion beforehand. In this paper, we explore the use of a low-cost wearable device equipped with two inertial measurement units (IMU) for learning reaching motion for real-time applications of Human-Robot Collaboration (HRC). A wearable device can replace or be complementary to visual perception in cases of bad lighting or occlusions in a cluttered environment. We first train a neural-network model to estimate the current location of the arm. Then, we propose a novel model based on a recurrent neural-network to predict the future target of the human arm during motion in real-time. Early prediction of the target grants the robot with sufficient time to plan and initiate motion during the motion of the human. The accuracies of the models are analyzed concerning the features included in the motion representation. Through experiments and real demonstrations with a robotic arm, we show that sufficient accuracy is achieved for feasible HRC without any visual perception. Once trained, the system can be deployed in various spaces with no additional effort. The models exhibit high accuracy for various initial poses of the human arm. Moreover, the trained models are shown to provide high success rates with additional human participants not included in the model training.

show abstract

“…Bamani et al [25] designed a wearable device for the user's hand, which was able to record data when identifying various objects by geometry. The device used Force Myography (FMG) sensors and was able to read muscle disorders in two areas of the hand.…”

Section: Prototypingmentioning

confidence: 99%

A Review on Wearable Product Design and Applications

Minaoglou,

Efkolidis,

Manavis

et al. 2024

Machines

View full text Add to dashboard Cite

In recent years, the rapid advancement of technology has caused an increase in the development of wearable products. These are portable devices that can be worn by people. The main goal of these products is to improve the quality of life as they focus on the safety, assistance and entertainment of their users. The introduction of many new technologies has allowed these products to evolve into many different fields with multiple uses. The way in which the design of wearable products/devices is approached requires the study and recording of multiple factors so that the final device is functional and efficient for its user. The current research presents an in-depth overview of research studies dealing with the development, design and manufacturing of wearable products/devices and applications/systems in general. More specifically, in this review, a comprehensive classification of wearable products/devices in various sectors and applications was carried out, resulting in the creation of eight different categories. A total of 161 studies from the last 13 years were analyzed and commented on. The findings of this review show that the use of new technologies such as 3D scanning and 3D printing are essential tools for the development of wearable products. In addition, many studies observed the use of various sensors through which multiple signals and data could be recorded. Finally, through the eight categories that the research studies were divided into, two main conclusions emerged. The first conclusion is that 3D printing is a method that was used the most in research. The second conclusion is that most research directions concern the safety of users by using sensors and recording anthropometric dimensions.

show abstract

Robust Multi-User In-Hand Object Recognition in Human-Robot Collaboration Using a Wearable Force-Myography Device

Cited by 8 publications

References 75 publications

Learning Human-Arm Reaching Motion Using a Wearable Device in Human–Robot Collaboration

Learning Human-Arm Reaching Motion Using a Wearable Device in Human–Robot Collaboration

Robust Classification of Grasped Objects in Intuitive Human-Robot Collaboration Using a Wearable Force-Myography Device

A Review on Wearable Product Design and Applications

Contact Info

Product

Resources

About