O. Arda Vanli

Florida A&M University - Florida State University College of Engineering

Papers

Total Citations

H-Index

About

O. Arda Vanli is a researcher whose work lies at the intersection of robust statistical estimation, sensor fusion, and Bayesian inference for dynamic systems. His primary contributions focus on solving the critical problem of unknown or misspecified uncertainty in state estimation, particularly in challenging environments. Vanli’s most notable work introduces a robust error estimation method based on factor-graph models for non-line-of-sight (NLOS) localization. This approach, detailed in his 2022 paper (9 citations), cleverly combines covariance estimation with M-estimators to mitigate the severe biases introduced by NLOS conditions, offering a general solution for linear regression problems. His foundational 2020 paper (7 citations) further advances the field by developing covariance estimation techniques for factor-graph-based Bayesian estimation, directly addressing the common yet invalid assumption that measurement and dynamics uncertainty is known a priori. By providing methods to learn these uncertainties online, Vanli’s work significantly enhances the accuracy and reliability of localization and tracking systems. His research is pivotal for students and engineers working on autonomous navigation, robotics, and wireless sensor networks, where robust performance under real-world, non-ideal conditions is paramount.