Testing / QA

In parallel to the development of our core products, we have developed a sophisticated testing methodology and platform that allows us to understand the capabilities and accuracy of the products in fine grain detail. The results of these tests allow us to refine the development of the traffic models we use, calibrate the models to enhance performance, and provide valuable information to our customers as to how and where the products deliver real benefit.

We have two main methods of testing we call  “End to end journey time testing” and “Link level analysis”. Both methods use data from hundreds of thousands of kilometres of real driven test routes, and are based on comparisons between actual speeds achieved during a test journey and forecasts based on our products according to the time of day, day of the week, vehicle type and individual driver. The end to end journey time tests compare the overall journey time forecasts against the actual journey times, whereas the link level analysis compares actual speeds achieved on individual road links in a test journey (the smallest element of the digital map) against the forecast link speed.

The journey time test results are split up into several categories relating to location (urban, suburban and rural) as well as time (peak, off-peak and night) to help show how consistent the results are in these different circumstances. We have found that the journey time tests give a strong indication of the overall user experience and system performance but do not readily expose the fine grain information required to develop or calibrate our traffic model – for this we use link level analysis.

In link level analysis we use the same real driven test route data but look at the complete speed profile across the entire trip making comparisons between actual and forecast speeds on every link. The results are cross referenced with information such as road type, setting, location, historic traffic events and other external databases, to give us a full picture of the performance across multiple aspects of influence.

These fine grain link level results and our journey time testing give us all the information we need to continue to develop our products and maintain our lead in terms of quality and consistency.