At a glance
- DatesMarch 2013 – March 2017 (four years)
- SponsorEngineering and Physical Sciences Research Council (EPSRC)
- Funded£929,000
- PartnersNetwork Rail, Schlumberger, BAE Systems, Sellafield Ltd, National Nuclear Laboratory (NNL), Scisys, Dstl (Defence Science and Technology Laboratory), the UK Space Agency, and the European Space Agency – the Autonomous and Intelligent Systems (AIS) partnership
This research project – AUTONOM: Integrated through-life support for high-value systems – is extending research in novel sensing, e-maintenance systems, and decision-making strategies.
Planning for maintenance in complex systems is difficult, expensive and uncertain. Degradation on the railway network can cause lateness, damage and hazards but regular measurement can lead to prevention, rather than cure. Automatic scheduling of the most cost-effective actions can send people to repair at the right place at the right time.
Our challenges include improvement of embedded sensing, reliable estimation of monitoring parameters, a unified approach to the mathematics and data structures, and a rigorous approach to cost estimation and benefit analysis.
Some of the industrial drivers include location, automation, connectivity, and optimisation of cost, over thousands of interventions on the network each year.
Progress update
Following an industrial workshop to identify needs, an integration strategy was created to bring together:
- Improved estimation of location, coupled with assessment of infrastructure measurements about condition of the railway track.
- Planning of the required actions to avoid damage and hazards, through early interventions devised from ‘big data’ searches with genetic algorithms.
- Activity-based costing, automated parametrically to deliver prioritised choice of planning.
A computer-based demonstrator has communicated the outcomes to a wide range of stakeholders across the Autonomous and Intelligent Systems (AIS) partnership, especially in Network Rail, and work continues to examine the architecture of unsupervised train-based track measurement.