Rudolph van der Merwe Senior Research Associate Adaptive Systems Laboratory Department of Computer Science & Electrical Engineering OGI School of Science & Engineering Oregon Health & Science University Email    :   r v d m e r w e @ i e e e . o r g Office : 150L, Bronson Creek Building, 1st floor

Current Employment

I left OHSU at the end of 2005 and worked for the South African Karoo Array Telescope (a Square Kilometer Array pathfinder instrument) project until September 2007.

In December 2007 I joined Apple and am currently a member of Richard Crandall's Advanced Computation Group (ACG).

Education

Ph.D : Electrical and Computer Engineering
    OGI School of Science & Engineering - Oregon Health & Science University (Portland, Oregon, USA), 2004.

M.Eng : Electronic Engineering
    University of Stellenbosch (Stellenbosch, South Africa), 1998.

B.Eng : Electrical and Electronic Engineering
    University of Stellenbosch (Stellenbosch, South Africa), 1995.

Research & Technical Interests

Probabilistic inference & machine learning. Dynamic Bayesian Networks. Data assimilation in large scale highly nonlinear dynamic systems. Neural, adaptive, and machine learning approaches to statistical signal processing, stochastic dynamic modeling, nonlinear estimation, pattern recognition, artificial intelligence, robotics and autonomous control.

Current & Past Research Projects

• Advanced Bayesian Estimation Techniques for Heterogenious Data Assimilation in Environmental Observation and Forecasting System (EOFS)

I'm currently part of multi-person team developing a state-of-the-art data assimilation system for the CORIE environmental observation and forecasting system (EOFS). CORIE is a pilot EOFS for the Columbia River system (river, estuary and plume), which integrates a real-time sensor network, a data management system and advanced numerical models.




• Sigma-Point Kalman Filters for Probabilistic Inference in Dynamic State-Space Models

Probabilistic inference is the problem of estimating the hidden variables (states or parameters) of a system in an optimal and consistent fashion as a set of noisy or incomplete observations of the system becomes available online. The optimal solution to this problem is given by the recursive Bayesian estimation algorithm which recursively updates the posterior density of the system state as new observations arrive. This posterior density constitutes the complete solution to the probabilistic inference problem, and allows us to calculate any "optimal" estimate of the state. Unfortunately, for most real-world problems, the optimal Bayesian recursion is intractable and approximate solutions must be used. Within the space of approximate solutions, the extended Kalman filter (EKF) has become one of the most widely used algorithms with applications in state, parameter and dual estimation. Unfortunately, the EKF is based on a sub-optimal implementation of the recursive Bayesian estimation framework applied to Gaussian random variables. This can seriously affect the accuracy or even lead to divergence of any inference system that is based on the EKF or that uses the EKF as a component part. Recently a number of related novel, more accurate and theoretically better motivated algorithmic alternatives to the EKF have surfaced in the literature, with specific application to state estimation for automatic control. We have extended these algorithms, all based on derivativeless deterministic sampling based approximations of the relevant Gaussian statistics, to a family of algorithms called Sigma-Point Kalman Filters (SPKF). Furthermore, we successfully expanded the use of this group of algorithms (SPKFs) within the general field of probabilistic inference and machine learning, both as stand-alone filters and as subcomponents of more powerful sequential Monte Carlo methods (particle filters). We have consistently shown that there are large performance benefits to be gained by applying Sigma-Point Kalman filters to areas where EKFs have been used as the de facto standard in the past, as well as in new areas where the use of the EKF is impossible.

See my publications for more detail.




• Sigma-Point Kalman Filter Based Sensor Integration, Estimation and System Identification for Enhanced UAV Situational Awareness & Control

As part of my Ph.D research I'm working on applying my SPKF based estimators and inference algorithms to UAV guidance, navigation and control. Here is the main project website.

Software

• A Matlab toolkit called ReBEL is available which contains most of my research code (or will eventually when all the bugs are worked out). For more information, go to the ReBEL page.

Miscellaneous Stuff and Links of Interest

1. Greg Welch and Gary Bishop's excellent Kalman Filtering web-site. This site has a wealth of information on Kalman filters, including an excellent introduction to the field. If you're new to the field, this is the place to start.