SOC Scientific Programmer
NASA Ames Research Center
I grew up in Germany and studied Physics at the University of Muenster, where I spent the last year of my studies working on the development of a volume holographic data storage system. After receiving my degree in 2003, I joined the Department for Theoretical and Computational Biophysics at the Max-Planck-Institute BPC in Goettingen to pursue my PhD. My research in protein folding mainly employed molecular dynamics simulations, and focussed in particular on the question how denaturants like urea change the energy landscape and induce protein unfolding. After finishing my PhD in 2007, I continued to work on related questions at the Max-Planck-Institute for another year as Postdoctoral Researcher.
In early 2009, I moved to California to take a position as Postdoctoral Researcher at Stanford University. Here I worked on the design of new protein mutants with improved functionality as compared to nature's wildtypes. This project was part of the NIH Nanomedicine Initiative and was particularly aiming to fight misfolding-related diseases like Alzheimer's, CJD, and certain types of cancer. As a side project, I also worked on algorithms for faster structure calculation to bridge the gap between computational models and experimental measurements.
It was also during that time that I decided to pursue my long-term interests in image processing, machine learning, and computer vision on a professional basis. As a result, I developed the tracking software AnTracks, which is now being used by insect researchers at various institutions for their studies on insect behavior and interactions, and colony organization. I also have experience on smaller projects in the areas of object classification and face recognition.
Outside of work, I enjoy spending time with photography, scuba diving, and classical piano. For more information, please visit my personal website.
Why I joined Kepler
I have a strong background in basic research, but to the same extent I enjoy developing methods and algorithms which find their way into real-world applications. Working on NASA's Kepler mission is a great way to pursue this passion. Here I can apply my expertise in image processing, machine learning and data analysis to contribute to a mission which scratches on the boundaries of our knowledge and has the potential to change our view of our place in the universe. Further, I think using sophisticated image and signal processing approaches to advance our knowledge about stars and planets which are thousands of light years away and seemingly completely beyond our reach is a very cool thing to do.
Role in Kepler
Before any planet transit signals can actually be detected in the light curve of a star, there is a substantial amount of signal processing required to "pre-condition" the data - which essentially means to remove all kinds of noise, systematic and random, while at the same time preserving the underlying astronomically relevant components in the signals. I am working on a Bayesian framework to perform this signal decomposition, and on other machine learning approaches to further improve the performance of the signal versus noise classification.