MSCA Doctoral Network AERIALIST

Introduction

In an era marked by groundbreaking innovations within assistive health technologies, people with motor disabilities have the prospect of achieving better health outcomes, enhanced treatment, improved quality of life, and increased participation in daily activities. This new way of healthcare has far-reaching implications, touching the lives of a diverse spectrum of individuals, from the expanding elderly population, two out of three requiring at least one assistive product to patients and healthy individuals navigating the challenges of automated work environments. Noteworthy examples include deploying rehabilitation robotics for home- based treatment, electric wheelchairs facilitating independent mobility, telehealth applications for cardiovascular monitoring, and powered prostheses restoring mobility and confidence in various daily life activities.

Current Landscape: The Promises and Limitations

While recent advances in hardware, sensing, and actuation technologies for assistive health technology showcase the immense potential for pervasive improvements in health and quality of life, significant hurdles remain. Cutting-edge systems face challenges when thrust into the uncertainties of real-world domestic environments, lacking adaptability and being burdened by the need for manual adjustments. Moreover, task-specific finetuning, safety, and reliability are not guaranteed in changing environments, and unforeseen circumstances can pose additional challenges. Good assistance is usually limited to supervised clinical settings, remaining far from the reach of individuals. The consequence is suboptimal assistance that is barely sufficient for most standard use-cases, at best, often leading to misuse or even the non-use of the assistive devices.

Urgent Call for Progress: Facing Reality

Statistical investigations in the UK and across Europe, show a surge in hospital admissions for acquired injuries and a concerning prevalence of amputations. This emphasises the critical need for advances in prosthetic technology and assistive healthcare. For example, Wheelchair wheelchair users face a significant challenge, with 87% reporting frequent falls that often result in traumatic brain injuries, highlighting difficulties to adapt to daily-life changes and challenging environments. This struggle within assistive technologies is emphasised by a rejection rate of 39%, 53%, and 50%4 for myoelectric hands, passive hands, and body-powered hooks among prosthetics users, mainly due to their limited functionalities and a lack of sensory feedback. Consequently, individuals using these prosthetics experience severe functional disabilities, hindering their daily activities. Moreover, major risk factors such as diabetes and peripheral artery disease, make these problems even harder to solve. For instance, Austria reported 11.4 major lower-extremity amputations per 100,0005 population with diabetes in 2019. Monitoring health through biosignals analysis and appropriate feedback can be crucial in addressing these issues.

In the realm of, industrial exoskeletons acceptance is hindered by issues such as cumbersome tuning, ease of use, compatibility with frequently changing tasks, and task-specific safety concerns. This reluctance6 has led to missed opportunities for overcoming longstanding, work-related muscle disorders in 60% of Europeans.

Addressing the major challenges that lead to today’s suboptimal assistance is vital to enhance the well-being and functionality of individuals with mobility and healthrelated concerns. Specifically, progress is crucial in health monitoring, ambient intelligent interaction, and regaining motor function.

AERIALIST’s Mission: Bridging the Gap

In response, AERIALIST takes on the challenge of advancing assistive health technologies to provide adaptive, reliable, and intelligent solutions. Our mission is to create systems that seamlessly adapt to users, environments, and tasks, offering human- like, intuitive, and symbiotic assistance. To achieve this, our research leverages the vast potential of recent theoretical machine-learning developments, translating them into real-world physical interactions between humans and engineering systems. The Cutting Edge: Realising AI for Real-World Physical Interactions As noted by [Zador et al. 20237], despite the remarkable strides in AI in-silico, such as ChatGTP, today’s AI systems fall short in comparison to the sensorimotor capabilities of a 4-year-old child or simple animals when applied to real-world physical interactions. AERIALIST stands at the forefront of artificial intelligence (AI) for real-world physical interactions, selecting one of the most challenging application fields: assistive health technologies. Here, the convergence of uncertainties in humans, engineering systems, diseases, and health problems presents an unparalleled opportunity for impactful innovation.

The Collaborative Journey Ahead: Merging Realms for True Equity

AERIALIST utilises a novel, cross-disciplinary approach to build and develop systems with real- world intelligence. Through collaboration with emerging doctoral candidates, experienced medical researchers, and innovative industry partners, we aim to merge the realms of sensing, control, safety, and learning.

Consortium: Beneficiaries and Associated Partners in the AERIALIST DN Project

This collaborative effort charts a new path for real- world physical AI, ensuring that assistive health technologies genuinely deliver on their promise— improving health outcomes, enabling accurate diagnosis and treatment, and ultimately enhancing the quality of life. With the aid of these devices, we want to transcend physical limitations, defying the very notion of what is possible, just as a real aerialist gracefully soars through the air. These assistive devices should become an extension of the body, offering support and the freedom to manifest their intention, forming a symbiotic relationship between the human and the technology. In this future vision, we are turning to the cutting-edge of AI and machine learning (ML), leveraging the remarkable strides. Our goal is clear: to provide true equity in the daily activities of those who need it most in the daily activities of those.

Objectives

WP1: Ensuring safe and reliable Assistive Health Technology

The AERIALIST project focuses on ensuring the safety and reliability of Assistive Health Technology. The growing integration of electronic controllers, digital sensors, ML algorithms, and wireless communications in compact and noisy spaces elevates the risk of disturbances from neighbouring systems, presenting significant challenges. A comprehensive safety framework is crucial to enable secure assistance, addressing all interconnected aspects seamlessly. Objective 1 aims to put a robust foundation for the development of safe and reliable assistive health technology in unsupervised home settings by ad- dressing intricate safety concerns.

WP2: Sensorimotor AI for assistive health technology

“Sensorimotor AI for Assistive Health Technology,” is a core pillar of the AERIALIST project. It focuses on utilising AI to bolster the sensorimotor capabilities of assistive health technology, enabling a deeper understanding of user needs. This involves creating intelligent perception mechanisms to enhance situational awareness and developing adaptive control strategies, including model predictive control, a hybrid approach combining physics-based modelling with data-driven techniques, and reinforcement learning. These innovations aim to make assistive technology more intuitive and responsive in unsupervised and home settings.

WP3: Symbiotic human Interaction for Assistive Health Technology

“Symbiotic Human Interaction for Assistive Health Technology,” aims to establish a seamless and harmonious interaction between humans and assistive health technology, cultivating an intuitive, responsive, and user-centric partnership. This involves a comprehensive analysis of well-synchronised interactions that are crucial for the success of assistive technology. The analysis actively promotes collaboration between biological and artificial nervous systems, heightens user interaction awareness, and facilitates the exchange of user-robot interactions across different assistive health technology systems. These endeavours ensure the development of genuinely symbiotic humanrobot interactions, especially during everyday tasks.

WP4: Application Case Studies to bridge the knowledge gap

“Application Case Studies to Bridge the Knowledge Gap,” is pivotal in the AERIALIST project. This objective employs an industry-driven methodology in four case studies to enhance the knowledge and understanding of assistive health techno- logy.
Encompassing critical areas such as (i) health monitoring and analysis, (ii) regaining motor function, (iii) independent mobility, and (iv) ambient intelligent interaction, (Figure 1.2d) these case studies aim to provide practical insights for developing robust, user-oriented assistive technology in unsupervised and home settings. This approach effectively bridges the knowledge gap by offering valuable data and insights that contribute to the progress of symbiotic assistive health technology.
Ready to join AERIALIST project?