Wednesday 21 December 2011 I defended my Ph.D. Thesis at Eindhoven University of Technology.
The digital version of my thesis can be downloaded here.

From 1 November 2011 I am working at Synerscope B.V. in Eindhoven.


Niels WillemsNiels Willems was a PhD student in the Visualization group of professor Jack van Wijk at Eindhoven University of Technology. His supervisor was Huub van de Wetering. We were involved in the Poseidon project of the Embedded Systems Institute, which investigated the system evolvability and reliability of systems of systems in favour of Thales Nederland bv and Noldus Information Technology bv. In this project, we focused on visualization of anomalies in maritime safety and security (MSS) systems.

Contact Niels C.M.E. Willems
SynerScope B.V.
Horsten 1, 5612 AX Eindhoven
The Netherlands

Email: email
View Niels Willems's profile on LinkedIn


Visualization of vessel trajectories for maritime safety and security systems

The sea seems an open playground, but regulations hold, and furthermore, safety (collision prevention) and security (threat prevention) need to be ensured. Operators monitor the coast using a Maritime Safety and Security (MSS) system, which allows analysis of multiple heterogeneous data sources. We aim at visualization methods to support operators to react on dangerous situations. To detect abnormal behavior, it has to be known what normal behavior is. As a first step, we start with visual exploration of trajectories of large vessels obtained by the AIS system.

Vessel density

Trajectories are convolved by moving a kernel with the speed of the vessel along the path. We compute the density with both a small and large kernel and show them simultaneously in the shading of a colored height field. The large kernel shows an overview of maritime highways and the small kernel reveals slow moving vessels, which typically occur in zones before the harbour where ships drop anchor. An interesting location is the entrance of Rotterdam harbour. On the left, earlier work of the Dutch Ministry of Transport is shown where ships observed from an airplane are convolved. The poster is constructed with about 8000 data points during 3 years. On the right, our new method shows millions of data points during one week of shipping movement. The image shows that captains nicely obey the maritime rules.

rws movement density

More info: Visualization of Vessel Movements, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk, Computer Graphics Forum (Proceedings of EuroVis 2009), vol. 28, no. 3, p. 959-966, 2009. size: 23 MB

Multi-source analysis

The data broadcasted with AIS may not be rich enough in the number of different attributes to find (spatio-temporal) patterns in trajectory data. By representing trajectories in a knowledge base, we are able to incorporate contextual information from the (semantic) web and use Prolog reasoning to annotate trajectories with higher level behavioral patterns. With these techniques we are able to detect ships that moor at harbours where they not supposed to moor. We do this by matching the ship type information from AIS and the web, and contextual information about the harbour from the web using Prolog rules. Artifical trajectories created with Presto enable us to verify the manually created Prolog rules. The annotations obtained from the web or by reasoning can be used next to the present knowledge to explore the attribute space with a multi-dimensional trajectory visualization.

Simple Event Model Multi-Source Visual Analysis

More info: An Integrated Approach for Visual Analysis of a Multi-Source Moving Objects Knowledge Base, Niels Willems, Willem Robert van Hage, Gerben de Vries, Jeroen H.M. Janssens, Véronique Malaisé, International Journal of Geographical Information Science, vol. 24, no. 10, p. 1543-1558, 2010.

Multivariate density maps

Temporal aggregation

Density maps are a generalization of vessel density, where attributes are taken into account in the computation and visualization of density fields. Multiple density fields are computed for parts of trajectories satisfying a filter criterium. The density fields are combined either into a new density field or the images of density fields are combined. Since the density fields are computed by means of graphics hardware, the density maps and their configurations can be explored interactively. In the publication, a number of relevant use cases are demonstrated showing the versality of our method, including temporal aggregation, anomaly detection, and risk assesment maps.

Anomlay detection Risk map

More info: Interactive Visualization of Multivariate Trajectory Data with Density Maps, Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk, Proceedings of PacificVis 2011, p. 147-154, 2011 size: 10 MB

Interactive visualization in large scale comparative genomics

MSA The genetic material of all organisms (DNA) is structurally and functionally equivalent. If two organisms share similar properties (phenotypes), these properties often derive from a common ancestor and are encoded similarly in the DNA. Comparative genomics is a molecular biological discipline searching for such similarities between DNA sequences as an indication of common ancestry and hence, a hint of similar function.

MSA Alignscope MSA vd Peer The number of similarities between DNA sequences may be enormous and chaotically distributed. Besides, on large scale, individual similarities might be negligible small, but can be part of a much larger similarity.
A large set of similarities between DNA sequences of Arabidopsis thaliana is hierachically clustered using various distance measurements to enable visualization using bundles. The result is an overview of dense locations of similarities, which current tools have problems to visualize. The overview of similarities can interactively be manipulated using the interactive DNAVis2 genome browser.

More info: Interactive Visualization in Large Scale Comparative Genomics, Niels Willems, Master's thesis, Eindhoven University of Technology, 2007. size: 15 MB





Color map: Journal, Conference, Workshop, Thesis, and Miscellaneous.
An extended list of activities.


Fellow visualization researchers

Poseidon researchers

Valid XHTML 1.0 Strict Valid CSS!