Grant: Discovery Grants Program - Individual
University of Alberta
$46,000.00
Continual Machine Learning for Intelligent Assistive Technology
EF 2024-25
Résumé
Building on the understanding and technologies generated during my past NSERC Discovery Grant on real-time machine intelligence for assistive technology, this proposed program of research will translate a new generation of methods from the fields of reinforcement learning and deep learning to the real-world application domain of upper-limb robotic neuroprostheses (artificial limbs for people with amputations).
1) As a specific target of this program, we will propose and refine new paradigms for continual machine learning during human-prosthesis interaction that support long-term, adaptive interactions between a person and their robotic limb. Approaches will be developed to enhance a) the control of a robotic prosthesis by its user, b) the utility and diversity of the feedback it delivers to its human user, and c) the predictive models a prosthesis builds of its user and its environment during use. A specific focus will be placed on machine-learned feedback, a largely under-explored area in prosthetics with substantial potential benefits to end users.
2) To drive the development of novel approaches to human-prosthesis interaction, my research program will create specialized robotic platforms alongside both physical and virtual-reality testing environments. These environments and platforms will allow us to fully investigate the implications that new continual learning approaches have for human users of a robotic limb with numerous modes, functions, and capabilities. Our new environments will also allow us to study a range of prosthetic software, hardware, and algorithmic configurations prior to their deployment and testing in long-term take-home trials by users.
3) Finally, for the first time, this program of research will evaluate the impact sustained continual learning has on neuroprosthetic control by humans in daily life settings and challenging competitive task settings. We will study co-adaptation between humans and prostheses using a suite of gold-standard quantitative and qualitative measures for assessing human sensorimotor decision making that were developed in our laboratory over the course of the last program of research.
The long-term impact of this work will be a transformation in how people and prostheses interact and how we study these interactions: never before have we been able to, as a species, build neuroprostheses that can continually improve and adapt alongside their human user during deployed use. My program will provide algorithms, environments, and evidence to realize this approach to human-prosthesis interaction in clinical and commercial design. In line with changing standards for rehabilitation medicine, it will also represent a profound step towards creating devices that respect the global diversity of prosthesis users: as users begin to define vectors of progress for their own rehabilitation, they can soon expect their intelligent assistive technology to change with them so as to achieve those goals.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Alberta, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2023-04450
Name nom
Pilarski, Patrick
Department département
Medicine - Medicine
Organization id
9
Institution établissement
University of Alberta
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
46000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1211
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Medical and health sciences
Area of application fr
Sciences médicales et de la santé
Research subject code
2800
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Artificial intelligence (computer vision, use2603)
Research subject fr
Intelligence artificielle (vision artificielle, utiliser 2603)
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Continual Machine Learning for Intelligent Assistive Technology
Keywords
reinforcement learning; machine learning; continual learning; real-time prediction learning; assistive rehabilitation technology; neuroprosthetic control and feedback; brain-body-machine interfaces; upper-limb robotic prostheses; human-machine interaction; multiagent communication and coordination
Application summary
Building on the understanding and technologies generated during my past NSERC Discovery Grant on real-time machine intelligence for assistive technology, this proposed program of research will translate a new generation of methods from the fields of reinforcement learning and deep learning to the real-world application domain of upper-limb robotic neuroprostheses (artificial limbs for people with amputations).
1) As a specific target of this program, we will propose and refine new paradigms for continual machine learning during human-prosthesis interaction that support long-term, adaptive interactions between a person and their robotic limb. Approaches will be developed to enhance a) the control of a robotic prosthesis by its user, b) the utility and diversity of the feedback it delivers to its human user, and c) the predictive models a prosthesis builds of its user and its environment during use. A specific focus will be placed on machine-learned feedback, a largely under-explored area in prosthetics with substantial potential benefits to end users.
2) To drive the development of novel approaches to human-prosthesis interaction, my research program will create specialized robotic platforms alongside both physical and virtual-reality testing environments. These environments and platforms will allow us to fully investigate the implications that new continual learning approaches have for human users of a robotic limb with numerous modes, functions, and capabilities. Our new environments will also allow us to study a range of prosthetic software, hardware, and algorithmic configurations prior to their deployment and testing in long-term take-home trials by users.
3) Finally, for the first time, this program of research will evaluate the impact sustained continual learning has on neuroprosthetic control by humans in daily life settings and challenging competitive task settings. We will study co-adaptation between humans and prostheses using a suite of gold-standard quantitative and qualitative measures for assessing human sensorimotor decision making that were developed in our laboratory over the course of the last program of research.
The long-term impact of this work will be a transformation in how people and prostheses interact and how we study these interactions: never before have we been able to, as a species, build neuroprostheses that can continually improve and adapt alongside their human user during deployed use. My program will provide algorithms, environments, and evidence to realize this approach to human-prosthesis interaction in clinical and commercial design. In line with changing standards for rehabilitation medicine, it will also represent a profound step towards creating devices that respect the global diversity of prosthesis users: as users begin to define vectors of progress for their own rehabilitation, they can soon expect their intelligent assistive technology to change with them so as to achieve those goals.