The technology programme

• 
  The requirement analysis will be correlated with the results of the CIC and UbiCare projects to identify technology gaps which will feed into initial service and system specifications. During the course of the project, the Gap analysis will be repeated using inputs from prototypes and evaluation of initial user trials to identify major areas for improvement of the system and means to effect such improvements.
  

No direct activity has been undertaken on A4 but related activities on understanding patient profiles and the role of a community matron have been undertaken as part of C1. These activities will feed in to Deliverable D06, “SAPHE prototype architecture analysis”, workshops to be held in August and September 2007 (delayed from June 2007). BT also met with two Community Matrons in South East Essex to solicit further insights into the community matron role and the potential value of SAPHE.

- variations

- reason

- recovery plan

• 
  IC - Ongoing research and analysis of existing and emerging technology solutions and commercial service offerings. Primarily this involves providing an understanding of needs, implementation challenges, and integration into current care system.
  
• 
  Docobo – Advisory inputs (when requested) into existing and emerging technology solutions.
  

• 
  Detailed technical and performance specification for the monitoring elements of the system and the ways in which they will interact. Detailed specification of the services required to deploy and manage large scale healthcare monitoring – includes communications, logging, training of personnel, etc.
  

Good progress has been made on defining a greater level of specification for the end-to-end SAPHE system in collaboration with Philips and Imperial. Work in this area has now moved to WPB6.

Continued provision of input for development top level specifications of body sensing networks and vision based monitoring systems. Meeting held between Philips, BT and Imperial to define specifications for end-to-end communications within SAPHE system (this also feeds into contributions towards WPB6), with particular input from Imperial regarding middleware.

- variations

- reason

- recovery plan

• 
  Imperial - Detailed specification of communication and middle ware based on UbiCare system experience.
  
• 
  BT - Development of Deliverable D07: Service Requirement for Large Scale deployment
  
• 
  Cardionetics - Development of wireless interface specifications ongoing
  

• 
  Design and development of common node architecture for SAPHE, including hardware and communications and operating software. Particular emphasis will be placed on low-power RF design, development of analogue STSOM neural networks, network self-configuration and security issues. The work package will also extend into the design and specification of WAN interfaces.
  

- variations

- reason

- recovery plan

• 
  BT - Integration and performance testing of ZigBee nodes
  
• 
  Imperial - Development of SAPHE node based upon extensions of UbiMon BSN
  
• 
  Philips - Complete low power radio; continue development of ultra low power radio
  

• 
  Use of common node technology for on-body sensing of activity and medical conditions. Autonomic sensing network with self-reconfiguration, self-healing, and context aware analogue implementation with STSOM neural networks.
  

- variations

- reason

- recovery plan

• 
  Imperial - Development of body sensing system. Identification and continuing work to incorporate greater range of sensors e-AR earpiece design. Development of middleware to support system integration and communications
  
• 
  Philips - Complete Synch software; integrate radio with sensors; optimise mobile hub design
  

• 
  Use of common node architecture for sensors around the home to determine activity and usage of facilities, e.g. motion, sound, vibration or temperature sensors. Emphasis on sensor message hopping for reduced power utilisation and network architecture that facilitates sensor inferencing, behaviour profiling, RFID applications, and support for self-configuration and self-healing.
  

Development of the ZigBee interfaces and network for environmental sensors has progressed well. A prototype ZigBee network has been programmed utilising the Integration DAMD1 modules upgraded with the latest Zigbee2006 firmware. Testing by both BT and Philips has identified range issues with the Integration USB dongles. This will be overcome by use of the DAMD1 modules (with a stub antenna) on the SAPHE hub (via serial port) in place of the dongles. This will also remove issues with a current lack of Linux support for the USB dongle. A data message protocol has been specified for the ZigBee network allowing a single sensor message to convey up to 83 bytes of sensor data. Support for fragmentation of longer data across multiple messages has also been provided.

Testing of the Tactex bed occupancy/pressure sensors has shown these to be of value for both simple occupancy detection and quality of sleep indication. An alternate commercial sleep analysis device available in Japan has been identified and it is hoped to obtain samples.
Imperial WPB4 – SAPHE Environmental Sensing

Development of PIR based motion sensor with ultra-low power radio for potential use within the SAPHE trial in order to minimise equipment costs. Ongoing enhancement of ‘blob’ sensors with focus on improvements for lower body tracking.

- variations

- reason

- recovery plan

• 
  Imperial - Integration of UbiSense/‘Blob’ sensor components
  
• 
  BT - Integration and performance testing of ZigBee nodes; Evaluation of third party sensors and integration methodologies.
  
• 
  Philips - Finalise software spec for set top box.
  

• 
  Population database design, data mining, behaviour profiling, information inferencing from recorded data and prediction of adverse effects.
  

Continuing progress and development of database functionality and access interfaces. Initial database architecture design complete. Document detailing SAPHE database structure and organisation of data along with XML format disseminated within the consortium. Work towards feature extraction, using multi-objective feature selection techniques, with focus on particular temporal features. Simulated datasets provided to Dundee for testing with appropriate visualisation techniques. WPB5 group teleconferences held in order to facilitate progress between partners and approach to initial development of analysis components agreed.

An approach to the initial development of analysis components has been agreed along side the architectural specification in B1. It has been agreed that each partner involved in data analysis (Imperial, Dundee and BT) will hold their own database. The prime requirement will be that raw sensor data is stored as-is. Partners will be free to implement their own database preference. Initially it will be left to the sensor ‘owners’ to create analysis components and visualisation tools for their sensor(s).

• 
  This has been our main area of effort over the past quarter.
  
• 
  We continue to use the ASP.NET with c# for development and the data being used is from the previous CitC trials.
  
• 
  The Dundas Chart software has been deemed to give the best results for the graphs required for the users, as well as having the capability present data help in OLAP data structures. However, because this is commercial software, it may be expensive to use in this project. For this reason, other graphing software has been investigated with some success. ZedGraph (http://zedgraph.org/wiki/index.php?title=Main_Page) is one which has shown to give good results in the early stages of using it.
  
• 
  Work in this area will continue as we look into other ways to represent data for the different end users. In particular, efforts will be made to produce prototypes that will work well on digital photo frames for informal carers.
  

- variations

- reason

- recovery plan

• 
  Imperial - Development of platform to integrate all data within dynamic system configurations. Particular focus on integrating sensor data from ambient and wearable sensors.
  
• 
  BT - Componentising of data analysis components form Care in the Community and Liverpool Pilot systems.
  
• 
  Dundee - Continue development for prototypes. Develop other ways of representing data, particularly the photo frame approach for informal carers.
  

• 
  Development of elements of service provision including communications systems, and implementation of interfaces between all users of the system. This will be based on the outputs of the initial service and patient engagement studies.
  

With progress on defining the end-to-end SAPHE system in collaboration with Philips and Imperial, a generic, standards based approach is to be adopted as much as possible for messaging between system components. A Java Messaging System (JMS) implementation will be used for communications between the home (Philips hub), a central SAPHE network platform (BT) and other distributed analysis platforms (Imperial and Dundee) each with their own database receiving raw data streams via a JMS broker. An XML based text format will be used for the messages. This will be a lightweight schema based on that in use by Imperial for the e-AR sensor. A VPN implementation for securing home to network connectivity has been tested and specified together with scripts for auto-configuration of a BT HomeHub (ADSL modem).

Work is underway on the technical specification document D07 ‘Service Requirement for large scale deployment’, this will build upon the previous documents pre-D07 and SAPHE Data Comms that have circulated between partners. A first draft will be circulated by BT in July ahead of the next quarterly meeting. This is a delay from June 06 due to delays in agreeing elements of the system architecture.
Imperial – Service Development

Development of user interfaces for inferencing system for service providers, based on research in WPA3 and WPB5 is underway.

- variations

- reason

- recovery plan

• 
  Imperial - Integration of data mining platform into service provision system. Development of user interface for inferencing system for service providers
  
• 
  BT - Development of task plan for this workpackage. Specification of Network platform connectivity with non-SAPHE systems, i.e. Liverpool PCT and BT Vital Life platforms.
  

[It is anticipated that this will be based on three mirror verification set-ups at IC, Philips and BT. The verification is based on the specification from B1]

• 
  Complete integration of communications, sensing and service elements of the system. Verification of performance against detailed performance specification, both at technical and usability level, and modification of design elements.
  

- variations

- reason

- recovery plan

• 
  Design and subsequent implementation of effective trials to establish the value of systems for use within an active care environment. NB: this will have required significant early work with healthcare provider partners to establish the test protocols, gain the necessary ethical approvals, and identify suitable patients or elderly users.
  

- variations

- reason

- recovery plan

• 
  Assessment of the value of the system in its various possible configurations to the several subpopulations within potential target user groups. Evaluation of business models for deployment of system to various target user groups. This will include both state organisations and private consumers.
  

- variations

- reason

- recovery plan

• 
  Management Group – weekly teleconferences and monthly progress meetings (ongoing)
  
• 
  WASP – EU common node project for wireless sensor networks (IC)
  
• 
  HaCIRIC – Health and Care Infrastructure Innovation (IC)
  
• 
  BiosensorNet – Novel biosensors and wireless networking technology (IC)
  
• 
  The Pervasive Computing Group have also recently won grant funding from EPSRC to develop Body Sensor Networks for Sports Applications (Sports-BSN), the grant will run from Jan 07 to the end of Jun 08 (IC)
  
• 
  Successful outline bid into MRC Lifelong Health & Wellbeing Centre proposal led by the Pervasive Sensing Centre
  
• 
  Visit to Nordic labs, Oslo, for outreach for SAPHE project, June 2007.
  
• 
  Co-ordination and submission of Q4 claims to DTI monitoring officer
  
• 
  Identification of Project Deliverables for publication outside of consortium (ongoing). This task will possibly take some time as Deliverables that have been identified will require some editing by their owners (Imperial, Philips and BT).
  
• 
  Next version of SAPHE Exploitation plan in development with inputs from consortium (led by management team) following attendance and feedback from DTI workshop on 19th June (Philips attended).
  
• 
  BEIC staff have linkages with the Platform for Assisted Living – launched 19 July 2007 in Birmingham
  

• 
  The second SAPHE Steering Committee Meeting took place on 6^th March 2007. The meeting covered project progress to date with a focused review of the scenarios and the SAPHE system in terms of maximising patient compliance. Also discussed were trial systems and costs, with consideration of security and trust issues, and future engagement with stakeholders. Notes from the meeting are available from: http://ubimon.doc.ic.ac.uk/saphe/public/SAPHE_SC_Mtg_6-03-07_notes.doc
  
• 
  Pervasive Computing Group at Imperial College have been preparing their interactive exhibit for display at the prestigious Royal Society Summer Exhibition 2 – 6 July 2007. The theme of the exhibit will be Body Sensor Networks, and is expected to attract high levels of public, academic and media attention.
  
• 
  Keynote presentation given by Prof. Yang at WOWMoM Conference on ‘From Theory to Practice in Wireless Sensor Networks 2007’, Helsinki, Finland.
  
• 
  Development of Pervasive Sensing Centre initiative at Imperial College, led by Prof. Yang and with cross-departmental collaboration, including: Department of Computing, Institute of Biomedical Engineering,
  

The SAPHE Project Spend Profile for Q4 Spend Summary shows some deviations from the baseline, these are:

• 
  Overspend on labour – this is largely due to a reduction in overhead rates for BT and Philips, which has enabled them to allocate more funds to labour costs (also reported in Q3 and Q4)
  
• 
  Underspend on overheads – see reason above and also reason below for Philips’ overspend on Subcontracts. Unfortunately, because the Project Spend Profile spreadsheet does not allow an ‘over-ride’ of the overhead rate after this has been entered into the baseline figures, a flag for incorrect overhead rate will show on the BT and Philips partner spreadsheets, even though the funds budgeted for overheads (rates have been reduced by both partners) can now be spent on labour costs.
  
• 
  Underspend on materials – this is a relatively small difference and is due to a reallocation of some materials costs by Philips to subcontracts, and the purchase of lab furniture by Imperial (see below)
  
• 
  Overspend on capital – this is due to the purchase of essential lab furniture by Imperial (reported in Q2 and Q3), not foreseen at the time of project planning, but required for the permanent trial environment at the IBE which has had to be reported under capital and not materials.
  
• 
  Overspend on subcontracts – this is due to internal staffing reallocations at Philips, which has meant that a number of Philips’ SAPHE project staff have had to be subcontracted from Philips APTEC (also reported in Q2 and Q3)
  
• 
  Overspend on other – see note for Philips below.
  

(L/M/H)

Note any issues regarding publications and confidentiality.