2 results
8 - Visual Analytics of Movement: A Rich Palette of Techniques to Enable Understanding
- from PART II - MOBILITY DATA UNDERSTANDING
-
- By N. Andrienko, Fraunhofer Institute IAIS, G. Andrienko, Fraunhofer Institute IAIS
- Edited by Chiara Renso, Stefano Spaccapietra, École Polytechnique Fédérale de Lausanne, Esteban Zimányi, Université Libre de Bruxelles
-
- Book:
- Mobility Data
- Published online:
- 05 October 2013
- Print publication:
- 14 October 2013, pp 149-173
-
- Chapter
- Export citation
-
Summary
Introduction
Visual analytics develops knowledge, methods, and technologies that exploit and combine the strengths of human and electronic data processing (Keim et al., 2008). Technically, visual analytics combines interactive visual techniques with algorithms for computational data analysis. The key role of the visual techniques is to enable and promote human understanding of the data and human reasoning about the data, which are necessary, in particular, for choosing appropriate computational methods and steering their work. Visual analytics approaches are applied to data and problems for which there are (yet) no purely automatic methods. By enabling human understanding, reasoning, and use of prior knowledge and experiences, visual analytics can help the analyst to find suitable methods for data analysis and problem solving, which, possibly, can later be fully or partly automated. In this way, visual analytics can drive the development and adaptation of computational analysis and learning algorithms.
Visualization is particularly essential for analyzing phenomena and processes unfolding in geographical space. Since the heterogeneity of the space and the variety of properties and relationships occurring in it cannot be adequately represented for fully automatic processing, exploration and analysis of geospatial data and the derivation of knowledge from it needs to rely upon the human analyst's sense of the space and place, tacit knowledge of their inherent properties and relationships, and space/place-related experiences. This applies, among others, to movement data.
12 - Air Traffic Analysis
- from PART III - MOBILITY APPLICATIONS
-
- By C. Hurter, France, G. Andrienko, Germany, N. Andrienko, Germany, R.H. Güting, FernUniversität in Hagen, M. Sakr, FernUniversität in Hagen
- Edited by Chiara Renso, Stefano Spaccapietra, École Polytechnique Fédérale de Lausanne, Esteban Zimányi, Université Libre de Bruxelles
-
- Book:
- Mobility Data
- Published online:
- 05 October 2013
- Print publication:
- 14 October 2013, pp 240-258
-
- Chapter
- Export citation
-
Summary
Introduction
The goal of air traffic control (ATC) is to maximize both safety and capacity, so as to accept all flights without compromising the life of the passengers or creating delays. Because air traffic is expected to double by 2030, new visualizations and analysis tools have to be developed to maintain and further improve the safety level. To do so, air traffic practitioners analyze data from the ATC activity. These multidimensional data include aircraft trajectories (3D location plus time), flight routes (ordered sequences of spatio-temporal points that represent planned routes), and meteorological data. In this chapter, we detail the relevant tasks of ATC practitioners and demonstrate recent visualization and query methods to fulfill them.
The special properties of ATC data propose new challenges and, at the same time, new opportunities of data analysis. The semantics of the data are rich because they includes the third dimension (altitude), which can be used to discover salient events such as takeoffs and landings. More semantics can be added by augmenting background data such as the traffic network and the meteorological data. ATC data sets are characterized by their large sizes, adding more challenges to the analysis. Trajectory analysis is difficult due to the data set size and to the fact that it contains many errors and uncertainties. One day's traffic over France contains about 20,000 trajectories (> 1 million records). Recording is done in a periodic manner (in our database: a radar plot, per aircraft, every 4 minutes), but a plot can be missed, or have erroneous data because of physical problems that occur at the time of recording.