ABSTRACT
he accumulation of a sufficient amount of data for training deep neural networks is a major hindrance in the application of deep learning in robotics. Acquiring real-world data requires considerable time and effort, yet it might still not capture the full range of potential environmental variations. The generation of new synthetic data based on existing training data has been enabled with the development of generative adversarial networks (GANs). In this paper, we introduce a training methodology based on GANs that utilizes a recurrent, LSTM-based architecture for intention recognition in robotics. The resulting networks predict the intention of the observed human or robot based on input RGB videos. They are trained in a semi-supervised manner, with the output classification networks predicting one of possible labels for the observed motion, while the recurrent generator networks produce fake RGB videos that are leveraged in the training process. We show that utilization of the generated data during the network training process increases the accuracy and generality of motion classification compared to using only real training data. The proposed method can be applied to a variety of dynamic tasks and different LSTM-based classification networks to supplement real data.
ABSTRACT
Occupational exoskeletons are not yet frequently used at work. Previous research has indicated multiple factors (physiological, implementation-related, work-related, policy, and psycho-social) that can explain this lack of adoption. However, there is a lack of specific requirements related to these themes to improve exoskeletons and their adoption. Therefore, the goal of our research is to formulate concrete requirements for the design and implementation of exoskeletons. We used a design ethnography approach (focus groups, an interview, observations, surveys, and group interviews) with multiple stakeholders (users, designers, and potential future users). Our data suggest that potential future users of exoskeletons believe that existing exoskeletons should be improved to be adopted. Exoskeletons should be more comfortable to wear and flexible enough to support a wider range of tasks. We have formulated 49 requirements for occupational exoskeletons, and we validated and extended an existing framework. Our work encourages the improved design of future or existing exoskeletons.
Fatma Demir, Ellen Roels, Seppe Terryn, Bram Vanderborght
ABSTRACT
Contribution: This article presents an engineering outreach activity that aims to teach K-12 students how to develop a tendon-based soft robotic finger. The primary objectives of this STEM activity are to introduce students to the fundamentals of soft robotics, its interdisciplinary nature, and to offer them a hands-on and engaging learning experience using the project-based- learning approach. Background: Soft robotics, an interdisciplinary field combining chemistry, materials science, and robotics, has the potential to revolutionize the design and development of robots. However, introducing the fundamental concepts of soft robotics to K-12 students can be challenging since traditional robotic activities often require complex programming, technical expertise and expensive equipment and software. Intended Outcomes: Increasing the students' understanding of soft robotics principles, materials, and polymer processing. Positively impacting students' perception of engineering as a potential career path by enhancing their attitudes toward STEM. Application Design: Students could develop manually actuated soft robotic fingers within a 45-min workshop by utilizing 3-D printed molds, rapidly curing elastomeric materials, and the basic mold casting method. The outreach activity is intentionally designed to simplify the technology used by eliminating the need for complex programming, and to focus on utilizing novel materials and basic concepts to construct actuating soft robots, providing an effective and engaging STEM activity for K-12 students. Findings: The success/effectiveness of the activity was evaluated in three ways: 1) through direct inspection on the performance of the student-fabricated soft finger during the workshop; 2) through the pre- and post-tests to evaluate the learning outcomes; and 3) by conducting a STEM outreach survey to gather student feedback on the quality of the outreach activity and their attitudes toward STEM. During the workshop activities, the students were able to effectively follow the instructions, construct a tendon-based soft robotic finger, and manually actuate the finger using the tendon. According to the results of pre- and post-tests, the students increased their understanding regarding the principles of soft robotics, materials and polymer processing. Furthermore, the STEM outreach survey of IEEE powered 'TryEngineering Portal' revealed that the developed outreach activity enhanced the achievement of pedagogical and quality outcome goals and measures, as well as program targets and objectives.