Research and Development: New Technologies of Care
We will create new data driven technologies to aid care and care research. Along with this, we will build a collaborative community that will provide insight for the development and evaluation of these technologies.
What are our intentions?
This research programme will develop practical, care-driven technologies that are fit for people in later life. In particular, this means exploring and developing data-driven IoT platforms that can produce accurate data about instant events (e.g. vital signs and serious incidents such as falls), short-term activities (e.g. those of daily living) and long-term pursuits (e.g. physical and mental activities over weeks and months) in order to extract (predictive) information and patterns that can be used, among other things, for effective interventions and the prevention of adverse outcomes.
We will create a community for collaboration that will provide the long-term context for the development and evaluation of multiple new technologies of care in a collaboration between academics from several disciplines, people in later life and their families, health and social care professionals, and businesses in this market. The collaboration will be formed around the implementation of routine monitoring in a highly vulnerable community in order to detect deterioration early, with a process to inform the focus of future work. We will both make a contribution to knowledge and practice, and support long-term development of new technical innovation.
For this purpose, we identify three main research goals:
- Implementing routine physiological monitoring in a home setting.
- Development and implementation of additional sensing methods and devices.
- Development of decision support and intervention management.
How will we achieve this?
The overall design of this work package is based on a set of 7 linked and complementary tasks that take advantage of an opportunity to generate sensors and technologies that complement care provided to people in later life. The research, which spans engineering and informatics, will use a variety of methods to produce platforms that combine elements such as multi-modal sensor integration, data analysis and fusion, AI-based decision and prediction modelling and assistive technologies to facilitate care for people in later life. For this work, we have established partnerships with a number of health and care partners who have agreed to provide expertise, data and other resources towards the development and validation of the outputs from the project. We will also work with colleagues in other work packages to prioritise the development of additional sensing parameters, AI-based decision support and assistive technologies, thereby ensuring we develop future-facing, innovative technologies that deliver the right care to people in later life.
Task 1: Comprehensive Review
The aim of this task is to carry out comprehensive reviews of the sensors and devices used for the monitoring of physiological parameters and activities and of the assistive technologies used for their care. These reviews will mainly inform the research being done in Tasks 2 and 7 but we expect them to be relevant to other tasks in different work packages.
Task 2: Devices and Sensing
This task will investigate the use of a number of available sensing methods to validate their effectiveness in capturing key physiological parameters, as well as, detect and predict the onset of falls through a number of key parameters. In the first instance, sensors will be embedded into fixed objects. Another approach would be to investigate integration of these sensors with personal items, such as bedsheets, clothes, patches, and other wearable items that are available in the care environment.
Task 3: Data Aquisition, Management and Fusion
In a continuous monitoring programme, it is important to process data online and in a timely manner in order to capture events such as falls or domestic accidents, that occur either unexpectedly or are triggered by a sequence of events, behaviours and external factors taking place over short intervals. This task is focused on maintaining a summary of the data collected via the multi-sensory platform in Task 2 in order to streamline their use in Tasks 4 and 5.
Task 4: Process, Performance and Predictive Modelling
This task will develop digitised models of care to allow for meaningful understanding and decision support. This includes models that capture knowledge about care pathways, guidelines, and practices, as well as individual needs and circumstances. These will be used to inform the new models of care work package. Explainable models using AI on existing data to produce alerts and insights towards meaningful interventions. Finally, we will use these digital models to provide predictive performance measures under different circumstances and ageing population trends.
Task 5: Behavior Modelling and Decision Making
This task will involve the creation of human-decision making models that can help with the workflows from Task 4. It will also build data-driven decision models that capture causal behaviour in order to provide coaching, e.g. helping the person better cope with exercise or medication regimes, and advice to the carer, regarding how activity variables are influencing measured outcomes.
Task 6: Infrastructure
In this task, we will build proof-of-concept systems for the different features developed across Tasks 2-5. More specifically, we will develop prototypes aimed at:
- IoT and data management
- The care receiver
- The care provider.
In a later phase, we will seek to use our prototypes as a testbed for our work in real environments. Issues of robustness, security, privacy, and always-on availability will be key considerations towards effective, AI-supported decision making and interaction at both the customer and carer ends.
Task 7: Intervention and Evaluation
This task will initially involve the development and evaluation of our technology as part of the Bayes Centre/School of Informatics ‘Living Lab’. It will involve a close collaboration with ACRC colleagues and our health and social care collaborators to leverage their understanding of the older person. This will ensure a user-centred, iterative design of our technology and of its evaluation. In a later phase, in collaboration with other work packages and our care partners, we will explore the possibility of setting up a real-world case-study to investigate a complex intervention and its evaluation.