FREE ENBIS Webinar by Dan Walker: "Data Science in the Improvement of Facilities Management Service Delivery"9 June 2020; 12:30 – 13:15
Dan Walker will talk about the data science in planned maintenance. The webinar will be moderated by Shirley Coleman.
Data Science in the Improvement of Facilities Management Service Delivery
Daniel Walker, Shirley Coleman, Jaume Bacardit
Often administrations choose to outsource the facilities management of public services to a specialist provider. Data science gives added value to the extensive datasets arising from maintenance issues. This case study is based on a contract with a local authority where the company provides planned and reactive maintenance for a range of assets in corporate buildings, hospitals and schools. Planned maintenance needs to be undertaken at fixed intervals and reactive work requires dynamic response to faults of varying urgency. Computer-aided facilities management (CAFM) systems are used to log jobs and monitor performance against key performance indicators (KPIs). Scheduling resources to ensure that contractual obligations are met in an efficient way is a difficult problem. The solution to this challenge involves natural language processing of job descriptions, exploratory data analysis and the design of bespoke job scheduling algorithms, which will be discussed and evaluated. We will also discuss issues around knowledge transfer from data scientist to operational staff and how change management processes were employed in this project to embed the new working practices.
Dan Walker Biography
Dan Walker is a Research Associate at Newcastle University working as a data scientist on a Knowledge Transfer Partnership with academic support from Dr Jaume Bacardit in the School of Computer Science and Dr Shirley Coleman in the School of Mathematics, Statistics and Physics. Dan was awarded a first class MMathStat degree from Newcastle University in 2018. He runs training workshops in R and Python for company staff and leads a coding club for local school children.