PROSEN (Networking of Distributed Sensors for Proactive Condition Monitoring of Wind Turbines) is funded by EPSRC (Engineering and Physical Sciences Research Council) from 1st October 2005 to 30th September 2008. This is a multi-University project, funded by EPSRC grants C014790 (Lancaster/Essex), C014804 (Stirling) and C547594 (Strathclyde). The project is also supported by external partners Agilent, Insensys, ITI Energy, Macom Technologies, NEL and Scottish Power.
Arrays of smart sensors are now routinely deployed for a wide range of applications in industry and government. These sensors are measurement transducers under the control of a local microcontroller that schedules measurements and preprocesses data. A typical example is the use of fibre optic strain gauges at wind farms.
This type of system is purely reactive, and prone to catastrophic failure in response to unanticipated failure modes, degradations, changed operating conditions or new environment conditions. The costs associated with these failures are prohibitive for remote sites such as offshore wind farms and pipelines, or for sites where frequent manual intervention cannot be justified such as coastal defence structures.
The research challenges being addressed by the project are:
- the design of systems with a rich self-observation capability
- the construction of machines that can continuously monitor parallel streams of input
- the development of effective, efficient and appropriate proactive control techniques
- the automated configuration and maintenance of large-scale sensor networks
- the specification and implementation of effective goal-driven configuration management
- the confederation of software services in such a distributed and changing environment
- the integration of novel research into a practical large-scale demonstration that encourages exploitation.
The work will have immediate benefits to users and providers of existing sensor arrays, and long-term research benefits for all users of management systems, autonomous systems, sensor and ad hoc networks.
The major objectives are:
- to demonstrate practical methods that have the potential to increase the mean time between failure of large-scale sensor arrays
- to facilitate re-configuration of the arrays to meet unexpected new needs
- to make fundamental contributions to the emerging field of proactive system management research.
The project is addressing these issues by investigating and demonstrating techniques that enable automated control and management of sensor arrays to be proactive. The project is facilitating the use of proactive control and management software through integration with:
- a policy-driven management infrastructure that uses high-level user goals to constrain the instrument-level proactive behaviours
- a platform that supports appropriate observation, communication and actuation.