Research Project

Data analytics and optimisation for predictive modelling of expenditure in automotive warranty

Principal Investigator
Professor Abdellah Salhi
A side-on shot of a row of cars, of various makes and models, parked on grass.

Predictive modelling of warranty expenditure in the automotive industry may be targeted, tailored to particular needs and circumstance, and made more accurate.

To these ends it can benefit from systematic modelling with a focus on optimisation, by exploiting current and historical data through analytics and machine learning, and fusing these with other accessible data relevant to the industry and attached to factors such as the road condition, the weather etc...

The data we use concerns products and dealers of these products, as well as third party participants such as the customers of these dealers.

Within the wider problem there are specific aspects that we examine in this project. Examples include:

  • Developing a predictive model of automotive warranty expenditure for a single model;
  • Generating advanced analytics and data visualisation to support a specific product that MSXi offers its customers;
  • Optimising the warranty auditing process, and so on.

These problems are a result of the limitations of the current system and approach to modelling warranty expenditure.

Throughout this project,  it is necessary to monitor and estimate the impact that the solutions may have on the MSXi as a business. We will also define methods for comparing financial and customer service performance and key performance indicators to measure success of both methodology and the tools we develop.

Partners

This project is run as a Knowledge Transfer Partnership, in collaboration with MSXi.

Funding

This project has received funding through Innovate UK.