Real-time Control of Upper Limb Exoskeleton: Active Torque Prediction Model Approach

Shyamendu Polis^*
*Correspondence: Shyamendu Polis, Department of Biomechanical Engineering, Mansoura University, Mansoura P.O. Box 35516, Egypt, Email:

Introduction

The advent of robotic exoskeletons in rehabilitation has opened new frontiers in the quest to enhance the functional capabilities of individuals with upper limb impairments. This study focuses on the real-time control of upper limb exoskeletons, employing an innovative approach centered on an active torque prediction model. As the demand for advanced assistive technologies continues to rise, the imperative to refine and optimize control strategies becomes evident. The incorporation of real-time active torque prediction not only improves the precision of exoskeleton movements but also paves the way for a more intuitive and responsive interaction between users and their robotic counterparts. This research holds promise in reshaping the landscape of upper limb rehabilitation, offering a glimpse into a future where robotic exoskeletons seamlessly integrate with human biomechanics to augment and restore lost functionalities. An active torque prediction model is a computational framework designed to forecast the torque, or rotational force, exerted by a system actively engaged in a specific task or operation. This model is particularly relevant in the context of dynamic systems such as robotics, machinery, or biomechanical systems where the prediction of torque is crucial for optimizing performance, enhancing control strategies and ensuring safety [1,2].

Description

The study is structured around the development and implementation of an active torque prediction model, designed to anticipate and adapt to the dynamic torque requirements during upper limb movements. Utilizing advanced sensing technologies, the exoskeleton continuously captures and processes real-time data related to joint angles, muscle activity and external forces. This wealth of information serves as input for the active torque prediction model, which, through machine learning algorithms and biomechanical insights, anticipates the torque demands associated with specific upper limb actions. The model's predictive capabilities are integrated into the real-time control system of the exoskeleton, facilitating instantaneous adjustments in torque delivery to align with the user's intended movements. A diverse group of participants, encompassing varying degrees of upper limb impairments, is engaged in the study. This inclusivity ensures that the active torque prediction model is robust and adaptable to the unique biomechanical profiles of individuals with different rehabilitation needs. The participants undergo a series of tasks that simulate daily activities, allowing for a comprehensive evaluation of the exoskeleton's performance in real-world scenarios. The research not only focuses on the technical aspects of torque prediction and control but also incorporates user feedback and subjective experiences, emphasizing the importance of usercentered design in the development of assistive technologies [3,4].

The development of an active torque prediction model involves the integration of various inputs, such as sensor data, system parameters and environmental conditions, to generate accurate and real-time predictions of torque outputs. Machine learning techniques, including neural networks, regression models, or physics-based models, may be employed to capture the complex relationships between input variables and torque generation. In robotics, for example, an active torque prediction model can enable precise control of robotic joints by anticipating the torque required for a specific movement or task. This predictive capability is essential for enhancing the efficiency and adaptability of robotic systems, leading to smoother and more accurate motion. In industrial machinery, the model can aid in predictive maintenance by anticipating changes in torque patterns that may indicate wear or faults in mechanical components. This proactive approach helps prevent unexpected breakdowns and reduces downtime, leading to increased operational reliability. In biomechanics and rehabilitation, an active torque prediction model can be applied to understand and optimize human movements, assisting in the design of assistive devices or rehabilitation protocols. By accurately predicting torque requirements during various activities, the model contributes to the development of personalized interventions tailored to individual needs [5].

Conclusion

In conclusion, the real-time control of upper limb exoskeletons through an active torque prediction model marks a significant advancement in the realm of rehabilitation robotics. By harnessing the power of predictive modelling and real-time adaptation, this research aims to address the intricacies of upper limb movements, providing users with a more natural and responsive interaction with their robotic counterparts. The potential implications extend beyond the immediate study, heralding a new era where robotic exoskeletons serve as seamless extensions of the human body, intuitively understanding and enhancing users' intentions. The integration of advanced control strategies, such as the active torque prediction model, represents a pivotal step toward the development of intelligent and user-centric robotic rehabilitation devices. If successful, this approach could redefine the standard of care for individuals with upper limb impairments, offering not just assistance but a transformative tool that actively collaborates with users to restore and augment their upper limb functionalities. As we navigate the intersection of robotics and rehabilitation, the real-time control of upper limb exoskeletons embodies the spirit of innovation and compassion, unlocking possibilities for enhanced independence and quality of life for individuals facing upper limb challenges.

Acknowledgment

None.

Conflict of Interest

No conflict of interest.

References

1. Gao, Baofeng, Chao Wei, Hongdao Ma and Shu Yang, et al. "Real-time evaluation of the signal processing of sEMG used in limb exoskeleton rehabilitation system." Appl Bionics Biomech 2018 (2018).

Google Scholar, Crossref, Indexed at

2. Song, Rong, Kai-yu Tong, Xiaoling Hu and Le Li. "Assistive control system using continuous myoelectric signal in robot-aided arm training for patients after stroke." IEEE Trans Neural Syst Rehabil Eng 16 (2008): 371-379.

Google Scholar, Crossref, Indexed at

3. Ghrouz, Amer, Ester Marco, Elena Muñoz-Redondo and Roser Boza, et al. "The effect of motor relearning on balance, mobility and performance of activities of daily living among post-stroke patients: Study protocol for a randomised controlled trial." Eur Stroke J 7 (2022): 76-84.

Google Scholar, Crossref, Indexed at

4. Bustamante Valles, Karla, Sandra Montes, Maria de Jesus Madrigal and Adan Burciaga, et al. "Technology-assisted stroke rehabilitation in Mexico: A pilot randomized trial comparing traditional therapy to circuit training in a Robot/technology-assisted therapy gym." J Neuroeng Rehabil 13 (2016): 1-15.

Google Scholar, Crossref, Indexed at

5. Liu, Yusha, Rafael ML Silva, Jeffrey B. Friedrich and Dennis S. Kao, et al. "Surface electromyography-driven therapeutic gaming for rehabilitation of upper extremity weakness: A pilot study." Plast Reconstr Surg 150 (2022): 125-131.

Google Scholar, Crossref, Indexed at