If you are interested in one of the topics listed below, please contact Prof Schömer.

In the automotive industry, the volume of a boot is calculated according to various standards (DIN 70020, SAE J1100). These standards define the volume of a boot by the maximum number of boxes that fit in the boot. A human packer needs a whole day for this task. An experienced packer manages to pack very well in this time. Our algorithms can solve this problem automatically. The quality achieved is comparable to or better than that achieved by hand. We use physics simulations and highly efficient geometric algorithms developed specifically for this problem. All this is embedded in a graphical tool that helps the user to interactively create high-quality packs.

We have developed algorithms for calculating the volume occupied by moving parts for construction planning and disassembly analysis.
One of our projects includes a method for determining the shape of the space available for moving machine parts. A specific task set by one of our industrial partners was to calculate the exact limitation of the volume filled by a moving motor. Such movements can be vibrations or caused by disassembly.

The video shows some applications of the curved volume calculation. One part shows the volume that a car engine takes up when it is removed from the engine compartment. Another application is the visualisation of the accessible space of a car seat.

The video shows a tool for real-time visualisation of engine displacement when dismantling an engine in an engine compartment. In this way, the structure of the volume can be recognised and interesting areas can be inspected. As no pre-calculations are required, we can manipulate the extended volume interactively by changing the movement path or the generator object.

Finding collision-free paths for the disassembly of parts and visualisation of tolerance violations.

Automatic calculation of collision-free (disassembly) paths. The video shows the disassembly of an engine in an engine compartment. In this way, digital models are used to determine whether parts of the final product can be removed for maintenance purposes. The second part of the video shows how we used our motion planning algorithm to solve a highly complex metal puzzle. Click to Play!

The video shows a tool for visualizing tolerance violations between two moving objects. As an example, the disassembly of an engine in an engine compartment is shown. All parts of the engine that infringes a certain tolerance distance are highlighted. It is then possible to inspect the entire path and identify difficult areas. The tolerance value and the movement of the motor can be manipulated interactively. In the second half of the video, we demonstrate the efficiency of our approach using two kite models with about 870,000 triangles each. The models are moved interactively by hand.

In order to do justice to the chaotic nature of weather phenomena in the field of weather forecasting, attempts are made to carry out simulations with many, slightly varied initial conditions. Based on the resulting ensemble of simulation results, one can assess the reliability of the forecast or estimate the probability of occurrence of certain weather events. In routine use, it is desirable to automatically identify and clearly visualize interesting weather phenomena in the ensemble. However, this requires that the essential structures of the weather phenomena can be sufficiently characterized and identified in large data sets.

In this interdisciplinary project, the meteorologists will further develop their methods for identifying and analyzing weather phenomena so that the computer scientists can design efficient algorithms to search the extensive simulation data for the weather conditions they are looking for. Minimizing external memory access and parallelizing the process will play a major role in this. After a careful object-oriented analysis of the problem, we will work on a meaningful and clear visualization.

Visit the INSIGHT-Website.

One of our algorithms generates dense arrays of circular parts in a circular container.
This can be applied to find dense packings of many cores in a larger composite cable. A high packing density minimizes the amount of jacketing material needed to wrap all the cores. Another advantage is the fully automated layout process, which eliminates the need to manually arrange all parts. Manual layout is often very time-consuming, as many boundary conditions have to be fulfilled simultaneously and the number of possible arrangements increases exponentially with the number of parts. A software solution also makes it possible to try out different scenarios and quickly change production layouts if requirements change.

In industrial production, the aim is to minimize waste when cutting out shapes from metal plates or cutting patterns from fabrics. To this end, we have developed algorithms to find dense arrangements of arbitrary shapes (non-convex polygons).