Skip to content

Traffic Incidents

Motivation

The visual analysis of traffic incidents is of high interest in the BOS-scenario. The real time gathering of traffic data and effective ways of visualizing aim to support emergency services in reacting to incidents or when searching for hot spots. Furthermore, the analysis of traffic and mobility data is of high interest for urban planning as it, for example, allows the planning of future infrastructure.

Visual Analysis of Traffic Incidents

Our visual analysis allows to gain overview as well as detailed representations of the underlaying data. At first glance, a calendar visualization of traffic incidents (Figure 1) allows to get an impression about the temporal distribution of the incidents. Figure 1, for example, allows the user to easily detect a repeating pattern in July 2016 where each monday for three weeks in a row was strongly noticeable.

Figure 1: Calendar visualization of traffic incidents.

Calendar Visualization

Additionally, we enable domain experts to inspect the categorical distribution of gathered traffic incidents. Domain experts can interactively decide for timespans of interest (e.g., only taking incidents into account which occur at night (Figure 2) or at afternoon (Figure 3)).

Figure 2: Categorical distribution of gathered traffic incidents occuring at night.

Barchart Night

Figure 3: Categorical distribution of gathered traffic incidents occuring at afternoon.

Barchart Afternoon

Ultimately, it is of high importance to provide adequate spatial visaulizations in order to allow domain experts to explore the data. We provide a more detailed animated heat map visualization on the map (see the video) as well as more abstract graph representation. The graph has been developed in order to detect patterns at larger scale. The graph is built hierarchically. On the top level, a spatial clustering is used to detect areas of interest. Afterwards, we split the clusters in the graph visualization based on their incident type by using color. On the next level, we indicate the severity by using saturation.

Figure 4: Several kinds of visual overviews to effectively aggregate the available data.

Graph and Heatmap View

Video

A video demonstrating some of the capabilities of the analysis of traffic incidents can be found by clicking below or here.

Visual Traffic Incident Analysis Demonstration