The Netherlands has a dense railway network infrastructure that needs constant maintenance in order to operate. This complex task is done by our customer, a government task organisation that takes care of the maintenance and extensions of the national railway network infrastructure. In order to do their long-term planning and budgeting, they needed a solution that would predict the remaining lifetime of various railway infrastructure components.
We helped the customer to improve short-term predictive models (max. 7 years ahead) and built long-term predictive models (7-30 years ahead). We performed data explorations of a plethora of data sources that were gathered throughout the organisation such as sensor data, ground penetrating radar data, ultrasonic data and much more.
An overall asset view that combined the developed models allowed for informed short-term and long-term decisions on maintenance and replacement planning. It also improved budgeting for the long term and helped them become more efficient in their core task as a government task organisation.