The Municipality of Amsterdam has multiple data sources that provide (real-time) insight of the traffic flows in the city. Three important data sources are FCD travel times, MoCo travel times from cameras and V-Log data from the traffic light systems (VRI). Every data source has its own limitation. No single data source provides the best information under all circumstances. Amsterdam wants to investigate whether the combination of these data sources can lead to a new data feed with better availability and quality. That was the reason why the municipality decided to conduct a pilot in order to investigate the potential of datafusion to generate an improved datafeed which can used for real-time traffic management and other Smart Mobility solutions.
Amsterdam invited 4 companies to implement the datafusion algorithm, including Scenwise. Scenwise accepted the challenge and has implemented the algorithm together with Bureau Onderweg and VORtech.
The algorithm was developed and validated on the s101 corridor with several traffic light systems. The algorithm is scalable to other trajectories in the cities. Amsterdam installed additional cameras on selected location in order to validate the outcome of the algorithm.
The project was conducted by 2 contenders in competition. We successfully implemented the datafusion algorithm. The outcomes of the pilot was evaluated by Amsterdam and we are proud to be the winner of the competition. Our algorithm delivers highly reliable outcome. Our method requires less configuration effort and is therefore scalable to the entire city or province. The datafusion application can work on historical data as well as real-time data.