📣 Help Shape the Future of UKRI's Gateway to Research (GtR)

We're improving UKRI's Gateway to Research and are seeking your input! If you would be interested in being interviewed about the improvements we're making and to have your say about how we can make GtR more user-friendly, impactful, and effective for the Research and Innovation community, please email gateway@ukri.org.

Empowering Augmented Mobility Of The Aging Population Using Computational Musculoskeletal Simulation

Lead Research Organisation: University of Bristol
Department Name: Engineering Mathematics and Technology

Abstract

Nearly 6 million elderly people nationwide, half of the aging population size, are physically inactive, with the consequence of ill-health conditions and decreased mobilities. The associated demand and supply of physical support care for the aging population have cost £5.3 million a year. Empowering mobility enables the aging population to preserve their physical fitness and promote health-related quality of life. Assistive robots have shown great potential for combating the mobility challenges experienced by the aging population. However, there are not currently unified consensuses about targeted joints and assistive strategies for assisting the elderly walking. Moreover, assistive robots fail to give direct clues on the effects of the user's efforts and energy changes, which renders experimental evaluations less informative in the design and evaluation.

The project will develop a computational simulation platform to evaluate the elderly walking with assistive robots and provide quantitative scientific evidence for optimal designs and customized assistive strategies. The proposed platform will simulate the elderly walking with assistive robots in a virtual environment by translating personalised characteristics of the aging population and models of assistive robots from the real world into digital forms. By incorporating real-world walking scenarios, engineering can exploit changes in muscular efforts and energy consumption to identify optimal robot design and customize assistive strategies. The proposed simulation platform provides a new solution to design and evaluate assistive robots for the elderly walking in an effective manner. It ultimately helps the aging population maintain mobility and increases healthy life expectancy.

Publications

10 25 50
 
Description From September 2023 to January 2024, multiple visits were made to a local retirement village to verify the user-centered assumption regarding the engagement of the aging population in weekly physical activities and also explored their needs concerning the use of an external device designed to assist them in walking more easily. By conducting a focus group in the local retirement village, the survey yielded several preliminary insights. The focus group comprised the most active elderly residents of the retirement village, who regularly participate in the exercise sessions provided by the facility. These findings contributed to the development of the 'User Persona' and the Value Proposition Canvas. Additional visits are planned to evaluate the retirement village's response to the newly refined product design.
In summary, the survey concludes:
1. The majority of participants prefer walking as their exercise to improve fitness, with more than half reporting concerns during walking.
2. Some elderly individuals use rollators or canes for support, expressing a need for aids that help maintain an upright posture to prevent hunching and lower back pain.
3. Falling is the most frequently mentioned concern among participants.
4. Participants have neutral feelings about the assistive device, recognizing a need for walking assistance during the discussion. However, they express a desire for independence in walking, suggesting that the device should offer resistance to aid in this.
5. The user-centered assumption is partially verified; the aging population is engaged in weekly walking exercises and has concerns during these activities. However, the initial identification of pain points regarding the walking assistive device and associated software needs further refinement through 'discover and define' methods."
Exploitation Route For the academic route, this research paves a new path for the design of high-tech powered devices aimed at enhancing the quality of life and improving healthy life expectancy among the aging population. It highlights the realistic challenges encountered by users of canes and rollators, such as hunching, lower back pain, fall prevention, and balance recovery. These issues can be transformed into engineering challenges and subsequently addressed by engineers. Furthermore, this research will continue the discover and define the pain point as well as iterate the product design, the outcome of the research data could be used to build up new products in the market to create non-academic impacts.
As mobility impairments become increasingly prevalent due to aging populations and rising rates of chronic conditions, intelligent walking assistance devices have emerged as a critical intervention addressing evolving user requirements. These sophisticated solutions integrate traditional mobility support with cutting-edge technologies-including artificial intelligence, Internet of Things connectivity, and advanced materials-to significantly enhance independence, safety, and quality of life for users. This analysis examines major geographical segments and consumer trends shaping this rapidly expanding sector. Both developed and emerging economies are experiencing substantial demographic transformations. The United Kingdom has witnessed an 18% growth in population aged 65 and above between 2015 and 2023. These demographic shifts have generated intensified demand for solutions addressing age-related mobility challenges. Contemporary users increasingly prioritise mobility aids that facilitate independence maintenance and confidence enhancement. Market research demonstrates heightened interest in devices enabling autonomous ambulation, balance improvement, and fall prevention. Technologies incorporating posture-correction systems, fall detection mechanisms, and real-time navigation capabilities are increasingly perceived as essential functionalities rather than supplementary features. The modern consumer conceptualises assistive devices as personal extensions, seeking solutions that balance functionality with aesthetic appeal and customisation options. Personalised design elements and integrated features such as wearable health monitoring capabilities contribute significantly to this trend, enabling simultaneous well-being management and individual expression.
The United Kingdom market specifically demonstrates a pronounced transition from basic ambulatory aids to intelligent solutions. Post-2020 adoption rates of IoT-enabled smart canes and wearable exosuits have increased by 32%. These advanced devices frequently feature integration with telehealth platforms, facilitating remote monitoring capabilities and enhanced rehabilitation outcomes. Rehabilitation centres and healthcare professionals have increasingly implemented powered mobility devices, including gait trainers and robotic exoskeletons. This adoption has been accelerated by National Health Service initiatives promoting early post-surgical mobility and reduced dependence on professional caregiving resources. With approximately 90% of care provision occurring within domestic environments in the UK, home setting integration has become a paramount consideration influencing product development. Lightweight materials and foldable architectures represent significant competitive advantages prioritised by end-users, reflecting the practical constraints of home-based mobility assistance.
[1] Arthur D. Little. (2023). The future of mobility 5.0. https://www.adlittle.com/en/insights/report/future-mobility-50
[2] Data Insights Market. (2023). Mobility aid medical device market consumption trends. https://www.datainsightsmarket.com/reports/mobility-aid-medical-device-1011126
Research and Markets. (2023). Personal mobility devices - Global strategic business report. https://www.researchandmarkets.com/reports/5303948/personal-mobility-devices-global-strategic
[3] Globe Newswire. (2022, February 24). Global assistive furniture market to reach US$5.3 billion by the year 2026. https://www.globenewswire.com/news-release/2022/02/24/2391279/0/en/Global-Assistive-Furniture-Market-to-Reach-US-5-3-Billion-by-the-Year-2026.html
Sectors Healthcare

Other
 
Title a DDPM-based signal reconstruction approach to restore corrupted signals encountered during HD-sEMG signal acquisition 
Description The applicant concluded their appointment at the University of Bristol in June 2024 due to personal family circumstances, which necessitated discontinuation of the project. Notwithstanding these constraints, the applicant completed research in the field of electromyography. These findings have been comprehensively documented and subsequently submitted for publication in a high-impact scientific journal. The manuscript is presently undergoing peer review evaluation by domain experts, representing a valuable academic contribution despite the project's conclusion. This work could be summarized as: High-density surface electromyography (HD-sEMG) has emerged as a powerful tool for advanced myoelectric control and muscle activation pattern analysis. The technology enables the collection of muscle activities through electrode array configurations, overcoming the limitations of bipolar electrode settings by assessing the spatial distribution of motor unit action potentials. However, HD-sEMG signal acquisition faces significant challenges attributed to its compact electrode array configuration, where individual channels can be affected by external disturbances including movement artifacts, power-line interference, and poor electrode contact. Conventional interpolation methods fail to effectively reconstruct corrupted signals, especially when multiple adjacent channels are affected. The performance of HD-sEMG based applications substantially relies on the quality and integrity of acquired signals. Signal corruption during experiments due to skin-electrode interface delamination and low signal quality can significantly degrade the utility of HD-sEMG signals for identification of muscle activation maps or precision of pattern recognition algorithms. The proposed approach addresses these critical challenges by offering a novel solution that reconstructs corrupted signals without requiring prior knowledge of corruption conditions, thereby enhancing the reliability and practical applications of HD-sEMG systems. This research introduces a novel HD-sEMG signal reconstruction approach based on the denoising diffusion probabilistic model (DDPM). The DDPM framework consists of forward and reverse processes that learn to capture the underlying distribution of HD-sEMG signals. Specifically, a U-Net network architecture combined with noising and denoising processes is developed to learn the data distribution. The method leverages both spatial and temporal embedding to capture the characteristics of HD-sEMG signals, reinforcing these spatiotemporal features during model predictions. To systematically evaluate the HD-sEMG signal reconstruction method, three distinct signal corruption scenarios were designed: random channel loss, random consecutive channels loss, and random regional channel loss. Each scenario incorporated four corruption ratios (12.5%, 25%, 37.5%, and 50%) and two temporal dimensions (6 model in total). The spatiotemporal channels were randomly emulated in all corruption scenarios to simulate realistic conditions encountered during HD-sEMG signal acquisition. The model was trained on both benchmark and self-collected datasets to demonstrate its generalization capability. The experimental evaluation revealed the superior performance of the proposed method over conventional interpolation approaches across various corruption scenarios for both benchmark and self-collected datasets. While interpolation methods demonstrated adequate performance at lower corruption ratios (12.5%), their effectiveness diminished significantly with increasing corruption levels. The proposed method consistently achieved lower normalized root-mean-square error (nRMSE) and higher peak signal-to-noise ratio (PSNR) across most modes and corrupted ratios. This research successfully developed a DDPM-based signal reconstruction approach to restore corrupted signals encountered during HD-sEMG signal acquisition. The U-Net network architecture, combined with the noising and denoising process, effectively learned the underlying data distribution without requiring prior knowledge of corruption conditions. Experimental evaluations on both benchmark and self-collected datasets demonstrated superior performance over conventional interpolation methods across various corruption scenarios with different levels. Despite its success, the study presents some limitations that warrant further investigation. Future research should focus on enhancing the network architecture to better handle severe temporal signal loss scenarios, developing generalizable paradigms for learning the HD-sEMG signal data distribution across subjects, and extending the framework for real-time processing applications. Nevertheless, the proposed method offers a promising solution for addressing challenges in HD-sEMG signal acquisition, potentially enhancing the reliability and applicability of HD-sEMG in prosthetic control and rehabilitation systems. Furthermore, More detail could be found in the future publication. 
Type Of Material Technology assay or reagent 
Year Produced 2024 
Provided To Others? No  
Impact This research successfully developed a DDPM-based signal reconstruction approach to restore corrupted signals encountered during HD-sEMG signal acquisition. The U-Net network architecture, combined with the noising and denoising process, effectively learned the underlying data distribution without requiring prior knowledge of corruption conditions. Experimental evaluations on both benchmark and self-collected datasets demonstrated superior performance over conventional interpolation methods across various corruption scenarios with different levels. Despite its success, the study presents some limitations that warrant further investigation. Future research should focus on enhancing the network architecture to better handle severe temporal signal loss scenarios, developing generalizable paradigms for learning the HD-sEMG signal data distribution across subjects, and extending the framework for real-time processing applications. Nevertheless, the proposed method offers a promising solution for addressing challenges in HD-sEMG signal acquisition, potentially enhancing the reliability and applicability of HD-sEMG in prosthetic control and rehabilitation systems. Furthermore, More detail could be found in the future publication.