Interactive Visualization of Air Traffic in OpenSpace

Abstract: This thesis report presents a master’s thesis project in Media Technology by two students from Linköping University, Sweden. The project was implemented in collaboration with the Visualization Center C and Linköping University during the spring of 2021 resulting in the creation and development of two spatiotemporal visualizations featuring air traffic data in the OPENSPACE software. One visualization uses live, real-time, data provided by The OpenSky Network through a Representational State Transfer Application Program-ming Interface (REST API). The other visualization uses a static historical data set covering aviation data during the COVID-19 pandemic that is mined from The OpenSky Network. A major challenge during the implementation was handling the large amount of data in a performant manner to avoid a reduced frame rate in the application. To solve this a mul-tithreaded method is used in order to not interrupt the rendering while new data are be-ing fetched to memory. OPENSPACE uses the Application Programming Interface (API) OpenGL to render graphics, thus also enabling a shader pipeline to be utilized. Multiple shaders are used to create the visualizations. The shaders’ purpose and implementation are explained in detail for both visualizations. The live data visualization features aircraft displayed by an anti-aliased trail of past po-sitions while the historical data visualization animates pathlines that represents aircraft with respect to time. The historical data visualization uses multiple Vertex Buffer Objects (VBOs) to render data efficiently and without interruption to allow the animation to play both forward and in reverse. The animation follows the time, time direction, and the speed of which the time plays set by the user in OPENSPACE to play seamlessly. The color of the pathlines are determined by continent to create a clustering effect without any pre-processing or calculations on the Central Processing Unit (CPU). Furthermore, the data can be filtered rapidly and in real-time by using Graphical User Interface (GUI) elements to control the filtering and by performing the actual filtering on the Graphics Processing Unit (GPU). The two visualizations enables user settings via their respective GUI. These settings include changing the color, opacity, and line width to aid in exploration of the data.