Direct cost reduction on system breakdown

Maintenance predictor

The municipal public transport operator in Amsterdam, GVB was under pressure to make use of its data to improve its service and to save on costs. Earlier PA had been analysing logging data from GVB trams and buses to identify vehicle stopping points. From these results PA identified opportunities to predict failure when other data would be added.

The challenge

How PA helped

Combining data from different parts of the organization, such as the occurrence of high-impact incidents and the duty roster of trams, we constructed a model that predicted brake failures with an accuracy tenfold that of the industry standard. Moreover, since we used factors applicable to all vehicles, our model could be used for the entire fleet, rather than just trams.

Benefit for the clients

P 950k Euro of annual savings potential identified by using data to predict & prevent incidents for its Combino trams. GVB can materialize quickly by starting with a small implementation & scaling up.