“The vision of SPHERE is not to develop fundamentally-new sensor technology specifically for individual disease conditions but rather to impact a range of healthcare needs simultaneously by employing data-fusion and pattern-recognition from a common platform of largely non-medical/environmental networked sensors in a home environment”
Work Packages | SPHERE House | Publications | Advisory Board
Read a feature article in the IEEE newsletter The Institute: 'The SPHERE House Can Monitor Its Residents’ Health'
The IRC work programme is divided into six Work Packages (WP1-6). In addition to the WPs, four transverse Activities (A1-4) will ensure close engagement with clinicians, patients and the public, in order that SPHERE technology is shown to be deployable in realistic scenarios and to achieve maximum impact. Through A4 all the IRC staff will work alongside clinicians to develop a deep understanding of clinical needs and practices.
Work Package 1: Environment Sensing
Lead Academic: Dr Rob Piechocki (University of Bristol)
With Dr George Oikonomou (University of Bristol)
The main project aim of WP1 is to develop a bespoke sensor platform for environment sensing and activity monitoring in a domestic setting. The system will collect structured data from multiple sensors. These will be used to build linked data sets that will enable clinical and non-clinical studies requiring home environment data, such as indirect activity recognition or investigating links between environment variables (e.g. temperature, humidity) to health monitoring and/or development, including onset, of medical conditions.
WP1 internal objectives are to produce (i) a common, expandable and interoperable ultra-low power wireless sensing platform that requires minimal end-user intervention or awareness, (ii) establish interoperable data structures and processes for accessing sensor data, (iii) investigate, jointly with WP6 and WP5, ontologies of home sensor data that maximise the opportunity for future-proof data analytical processing.
Work Package 2: Video Monitoring
Lead Academic: Professor Majid Mirmehdi (University of Bristol)
With: Dr Dima Damen (University of Bristol)
The main objectives of WP2 consist of developing an efficient, real-time multi-camera system for activity monitoring in the home environment. The system is based on low cost cameras and depth sensors to estimate client’s position and to analyse their movements to extract features for use for action understanding and activity recognition. SPHERE's particular challenge is to develop algorithms that work robustly for months in occluded, cluttered, badly-lit and multi-resident environments.
The Work Package will be built upon the implementation of a real time multi-camera system comprised of a combination of low-cost RGB-D cameras (Microsoft Kinect, Asus Xtion, Prime Sense) and wide angle traditional monocular low-cost colour cameras (such as Genius WideCam 1050). Each SPHERE home is equipped with 2-3 such devices, with the RGB-D cameras being stationed advantageously such that specific actions such as sitting to standing (in front of the television) and ascending the stairs might best be observed.
Work Package 3: Body area sensing and communications
Lead Academic: Dr Rob Piechocki (University of Bristol)
With Dr Geoff Hilton (University of Bristol), Professor William Harwin (University of Reading), Professor R Simon Sherratt (University of Reading), Dr George Oikonomou (University of Bristol)
The WP3 Vision is to develop robust, energy efficient and user acceptable wireless communication system and wearable devices. The system consists of a bespoke wrist-worn wearable that continuously streams data to am in-house network of access points. The wearable devices will be small and designed to function without user intervention - battery life and reliable connectivity are key drivers. The system will provide real time information regarding physical activity and localisation, including the ability to fuse those sources of data with the other SPHERE components.
Work Package 4: Energy harvesting and management
Lead Academic: Dr Bernard Stark (University of Bristol)
With Professor Steve Beeby (University of Southampton)
This work package combines the material science expertise at Southampton and the low-power systems expertise at Bristol to investigate methods for powering autonomous body-worn healthcare devices. Both Universities are investing a significant portion of the project resources into short to medium range inductive power transfer, as a viable route to supplying mW loads within the duration of the project. The energy harvesting and RF power transfer methods will power bespoke ultra-low power systems, which are heavily duty-cycled.
Work Package 5: Data fusion and data mining
Lead Academic: Professor Peter Flach (University of Bristol)
With Dr William Harwin (University of Reading)
The information from the individual sensors developed in WPs 1-4 must be pre-processed, integrated and mined to provide a most likely model of activity which maximises information content in a health monitoring context. Moreover, the decision-making process (e.g., distinguishing between regular activities and medical emergencies) needs to be implemented and fine-tuned taking into account the specific characteristics of sensors and people.
Work Package 6: Design, deployment and monitoring
Lead Academic: Dr George Oikonomou (University of Bristol)
WP6 will provide the SPHERE end-to-end system integration and deployment activities. The proposal text shapes WP6 into two main activity areas: (i) to deliver an integrated in-home sensor network that brings together outputs from home environment sensors (WP1), video sensors (WP2) and on-body sensors (WP3) and delivers their data using suitable secure communications to a central server data warehouse (SPHERE Hub) hosted by University of Bristol; (ii) and to deploy the system over a number of up to 100 homes in Bristol in Year 4 of the project (optionally, starting in Year 3). The second activity area is envisaged to be split into two, consecutively running, activities: deployment planning and user engagement using demonstrator platforms, and the full deployment and monitoring stage, originally envisaged from Y4 onwards.
Activity 2.1: Fall prediction
Lead Academic: Prof. Ann Ashburn (University of Southampton)
A leading example of clinical engagement is Activity A2.1, led by the Faculty of Health Sciences at Southampton, this activity is gathering initial data on Parkinson's Disease patients in their homes, in order to indentify behaviours associated with falling. Data collection from these patients, carefully supervised by a Health Science researcher, is employing both accelerometry and video. The research is helping to understand risk of falling with the aim of ultimately designing interventions in the later years of the project.
Lead Academics: Dr Rachael Gooberman-Hill (University of Bristol) and Dr David Coyle (University College Dublin)
The main aim of A3.1 is to contribute to the design of meaningful and desirable SPHERE technologies. If healthcare technologies are to become embedded into people’s everyday life, their development must be informed through early and sustained user involvement. The user-centred design framework therefore involves collaboration with a range of stakeholders, such as domestic users, care givers and healthcare professionals. The methodology seeks to use participatory techniques, which empower and engage these stakeholders in the research process.
SPHERE's PhD students
SPHERE has several PhD students working on the project on a range of subjects. SPHERE is a pioneering multidisciplinary project, providing excellent opportunities for PhD students to learn in an exciting and challenging real-life project.