Investigating the use of novel sensors for monitoring applications: to include the fusion of such sensors with more conventional solutions

Lead Research Organisation: University of Cambridge
Department Name: Engineering


Context of the research:
The development of innovative home monitoring systems for elderly people would allow for novel approaches to treat, monitor, and even diagnose elderly people in healthcare systems. In particular, two emerging categories have attracted increasing attention of researchers in recent years. Computer-vision based techniques and radar-based techniques have been investigated separately to provide home monitoring, as they do not encroach on users' privacy and do not require constant wearing of any additional devices.

This project aims to develop an indoor monitoring system for elderly people, integrating radar sensing and computer vision. The fusion of measurement data from radar sensors and computer vision has the potential to increase localisation accuracy for monitoring of the elderly in a home environment.

Moreover, this integrated system can provide crucial healthcare-related information, enabling medical or healthcare professional to make useful inferences on the health condition of those being monitored.

It is economically and socially advantageous to develop and implement coherent and lower cost systems to provide affordable healthcare and monitoring services aiming at the ageing population. If the evaluation of the system proves to be valid, accurate and stable, there could be a commercial opportunity to patent the intellectual property or to form a start-up company to commercialise the wireless integrated home monitoring system for elderly care applications.

Aims and objectives:
The project aims to demonstrate a radar-computer vision-based home monitoring system. Through the development of this platform, it could allow the integration of both sensing, recognizing and tracking of the target people, inferring target physiological signals by analysing RF signals using machine learning. Ideally, the platform has potential to review symptoms in neurological conditions, e.g. Parkinson's disease.

The project will be structured in two main parts with the following objectives:
- Development of the radar and computer vision platform triggering vision switch-on if the radar indicates. Both conventional and machine learning methods will be investigated for the extraction of features from the raw radar data. The database collected by the radar sensor will do further prediction to set conditions for camera triggering.

- Implementation of a radar triangulation system on the radar and computer vision integrated platform: this will involve new signal processing to deal with the combination of noisy radar data and investigation of radar chip placement determination. Backend signal processing electronics will also be developed to allow tracking features to be extracted.

Research Methodology:
An integrated system will be developed, which will enable us to create a database of target tracking data produced from the fusion of radar and computer vision systems. The novelty in this research project will lie in increased accuracy for localisation and tracking, through combining the radar data with vision. With the dataset collected, machine learning techniques will be applied in order to extract biometric parameters and indications of disease/impairment. Performance will be evaluated and compared with more conventional sensors. Through the development of this novel integrated home monitoring system for the elderly, we hope to produce a useful tool for the healthcare and clinical treatment sectors.

EPSRC's strategies and research areas:
This PhD project is related to the research themes of ICT, healthcare technologies and engineering.

Any companies or collaborators involved:
During this project, we aim to establish valuable collaborations between research group and company. We have experience dealing with signal processing, human motion detection and the machine learning model, but would benefit from company's knowledge and collaboration on sensor hardware.


10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/S023046/1 01/10/2019 31/03/2028
2262357 Studentship EP/S023046/1 01/10/2019 30/09/2023 Qian Wang