Using Data Science to Improve Asset Resilience and Environmental Performance

点击翻译按钮获取中文摘要

**Background**Utilities and other asset-rich sectors are under intense pressure from governments, regulators and the public to improve their environmental performance, and the operational performance, safety and reliability of their assets, for example pumps, whilst working under more intensive scrutiny and regulation than ever before. They require an asset management system that satisfies these demands and the additional demands posed by climate change by minimising the risk of asset failure. Our innovation is concerned with the development of a predictive asset management that minimises the risk of asset failure and therefore the risk of damage to the environment caused by asset failure.**The Environmental Challenge**Our innovation improves the environmental performance of organisations and reduces the energy used by their energy intensive infrastructure assets and so helps them reach net zero. The consequences of asset failure include flooding, pollution, service interruption, significantly higher costs, very large fines from regulators and low asset availability.The energy used by energy-intensive assets depends on how they are maintained. If assets are not in optimal condition, they use much more energy than expected and are at greater risk of failure. Assets whose asset management policy is based on their calculated current risks of failure rather than on a fixed frequency can be scheduled for proactive maintenance when they need it, i.e. when they are not operating optimally and have high risks of failure. Our innovation identifies the risk of failure of each asset as it is used, maintained. fail and then reinstated. Preliminary simulation results from the innovation show that the risk of asset failure is reduced significantly by only small increases in the level of proactive maintenance.**The Solution**Our innovation is a cloud hosted, secure, scalable predictive asset management product offered as SaaS. It optimises asset management by:\* modelling, simulating and optimising asset management at individual asset level and at the operational, tactical and strategic levels\* modelling each asset as a unique entity with its own dynamic risk of failure profile so that each asset can be monitored rather than treat each asset as a member of a cohort where all assets in a cohort have the same risk of failure profile\* having exploratory data analysis and data preparation functionality to improve the quality of the data and ensure that the data are in the correct form for analysis and modelling (the essential first step in all analytics projects).