Microscopic Traffic Simulation (SUMO / VISSIM)

Definition and Technical Concept

Microscopic traffic simulation represents the most granular tier of traffic simulation systems. Taking individual vehicles as the fundamental modelling unit, it realistically captures behavioural details such as car-following, overtaking, and lane changes on the road. If macroscopic simulation is akin to observing the overall "flow direction" of a river, microscopic simulation tracks the "movement trajectory of every single drop of water". It serves as the go-to tool for evaluating junction designs, traffic light timing, and autonomous driving algorithms, with SUMO and VISSIM being the two most prominent software suites in this domain.

SUMO: Open-Source, Flexible, Large-Scale

SUMO (Simulation of Urban MObility) is an open-source, highly portable microscopic traffic simulation software package developed by the German Aerospace Center and currently maintained by the Eclipse Foundation. Its open-source architecture makes it a widely adopted platform for academic research and the validation of autonomous driving algorithms. SUMO’s core capabilities encompass the explicit modelling of individual vehicle trajectories and interactive behaviours, high-efficiency processing of city-wide road networks, and multi-modal simulation support covering motor vehicles, pedestrians, and non-motorised transport. It supports the import of standard map formats like OpenStreetMap, features built-in road network editors alongside demand generation tools, and provides a TraCI interface enabling users to dynamically control vehicle behaviour during simulation runtime. However, SUMO's visualisations are comparatively rudimentary, relying predominantly on a 2D graphical interface, making it ideal for algorithmic testing rather than executive-level presentations.

VISSIM: Commercial, Detailed, High-Fidelity

VISSIM is a commercial microscopic traffic simulation suite developed by Germany’s PTV Group, first released in the 1980s. Regarded as the commercial benchmark of the industry, VISSIM is extensively utilised for modelling urban roads, expressways, and car parks, simulating diverse transport modes including passenger cars, lorries, buses, bicycles, trams, and pedestrians. VISSIM’s primary competitive edge stems from its sophisticated driver behaviour model; its car-following logic is built upon Wiedemann’s psycho-physical model, replicating human driver response dynamics with high fidelity. Furthermore, VISSIM features advanced 3D visualisation capabilities capable of generating realistic traffic environments complete with surrounding architecture and pedestrian flow, rendering it exceptionally well-suited for engineering project demonstrations and stakeholder briefings. It also supports seamless integration with GIS platforms and mathematical tools, alongside a dedicated traffic emissions modelling module co-developed with Bosch.

Comparison and Choice Between the Two

SUMO and VISSIM diverge across several key dimensions. On the licensing and cost front, SUMO is completely free and open-source, whereas VISSIM is a commercial product requiring proprietary licence fees. Regarding visualisation and presentation, SUMO relies mainly on a 2D interface with straightforward graphics, whilst VISSIM delivers high-fidelity 3D rendering tailored for project showcases. In research and development, SUMO sees broader adoption across academic studies involving autonomous driving algorithms, vehicle-to-everything (C2X/V2X) communications, and dynamic traffic assignment, favoured by research bodies for its open-source flexibility and programmatic control; conversely, VISSIM remains the staple across industry applications such as traffic engineering consultancy, urban transport planning, and project feasibility assessments. Notably, a comparative study led by the Oak Ridge National Laboratory revealed that across identical scenarios, network-level traffic performance metrics from VISSIM, SUMO, and another platform, AIMSUN, were broadly aligned, showing minor discrepancies in energy consumption metrics only under peak traffic demand conditions.

Technical Limitations

The efficacy and reliability of microscopic traffic simulation hinge heavily on the calibration precision of model parameters. If core parameters—such as time headway and driver reaction time within the car-following model—are not calibrated against real-world field data, simulation outputs can diverge significantly from actual road traffic conditions. Meanwhile, in the Chinese market, domestic microscopic simulation software such as TESS NG is gaining rapid traction, having already been deployed in major initiatives including the Beijing Winter Olympics digital twin platform and the Hangzhou Bay Bridge smart expressway traffic management system, offering the industry a viable alternative beyond conventional commercial software suites.