Data driven predictive maintenance optimisation for the intelligent management of assets

Since the 1970s there has been an increasing demand for integrated maintenance philosophies. As a result there has been an evolution of data driven maintenance calling for transition from time-based maintenance to proactive maintenance based on condition. However, the industrial application of such concepts appears to be restricted to industries such as Aviation or Formula 1. In the built environment the management of safety, reliability, risk and cost impacts are considered as highly as the before mentioned industries yet the successful application of proactive condition based theories and technologies to assist with such concerns appears to be extremely rare.

Therefore, this research focuses on the investigation of fault detection, diagnosis and prognosis on mechanical and electrical building assets via numerous sensors and systems for efficient and innovative asset management in order to address the problem of asset failure, reducing costly reactive maintenance efforts and enabling proactive informed decision-making. The principal focus is optimization through statistical analysis and machine learning of maintenance programmes based on data relating to key asset operating parameters, mechanical vibrations and the environmental conditions. Additionally, the relationship between the cost of implementing these technologies in practice against the return on investment, benefits and the associated business risk will also be explored.

People

PI:  Michael Pitt                                   

Co-I:  Peter McLennan  

RR: Ruhul Amin

Outputs

Planned outputs arising from this research include:

Research exploring maintenance strategies and understanding the impact of environmental conditions on fault inception and degradation. Practical industry based implementation of innovative techniques to optimise fault detection, diagnosis and prognosis. Framework for transition from scheduled maintenance to proactive Condition Based Maintenance enabling better asset condition data to inform Life Cycle decisions. Novel, intelligent maintenance model that has potential for wider industry application. Numerous publications in the areas of (i) asset condition monitoring and maintenance, (ii) FM supply chain management (paper accepted in journal of FM), (iii) strategic asset management and the impact of environmental conditions on asset life degradation. 

Impact

The potential main impact of this project is to initiate industry-changing developments in the field of maintenance management. 

Links

HCP

For further information please contact: Michael Pitt