The laboratory uses a system of multi-agent simulation of transport systems based on open source products (MATSim, SUMO, UrbanSim and others) that allows flexible use of heterogeneous input data and solve non-standard problems using advanced software functionality.
Multi-agent models are based on the description of individual city resident’s behavior during a whole day and allow to analyze transport parameters in relation to the overall system, a separate carrier, an individual vehicle or person.
Such models allow to unleash the full potential of using new data types – data on movement of cellular network subscribers, or validation of proximity transport cards – in transport planning.
- Creating a multi-agent model of Krasnoyarsk and the Krasnoyarsk agglomeration
- Creating global transport service models for the 2017 Confederations Cup and the 2018 World Cup
- Creating a multi-agent model of Moscow based on mobile data on citizens’ movements
- Creating a pilot multi-agent model of St. Petersburg as part of work on the concept of “smart city of St. Petersburg”