Download Call For Tutorials Click here to download the call for tutorials.

List of Tutorials of MFI 2016 T1 Introduction to distributed and event-based state estimation T2 Relatives for the Kalman filter for tracking and fusion T3 Robust Kalman filtering T4 Multisensor fusion in Automated driving T5 Fusion and Bayesian Reasoning with Subjective Logic T6 Passive Surveillance - Advanced Algorithms and Challenging Applications T7 Extended object tracking: Theory and applications T8 Proactive Optimal Control to Infer Information Faster

Time Table for the Tutorials Day on Monday, September 19, 2016 Room A Room B Room C Topic Tracking & Control Sensor MFI State Estimation Morning 08:00–11:00 T7 Extended Object Tracking: Theory and Applications K. Granström T4 Multisensor MFI in Automated driving J. Elfring T1 Introduction to distributed event-based state estimation S. Trimpe 11:00–11:30 Lunch break Mid Day 11:30–14:30 T8A Proactive Optimal Control to Infer Information Faster R. Urniezius T5 MFI and Bayesian Reasoning with Subjective Logic A. Josang T2 Relatives of the Kalman Filter for Tracking and MFI D. Fränken 14:30–15:00 Coffee/tea break Afternoon 15:00–18:00 T8B Proactive Optimal Control to Infer Information Faster R. Urniezius T6 Passive Surveillance - Advanced Algorithms and Challenging Applications W. Koch T3 Robust Kalman Filtering F. Pfaff 18:00–20:30 Welcome Reception

T1 Introduction to distributed and event-based state estimation Presenter: Sebastian Trimpe Length: 3 hours Brief description: This tutorial provides an introduction to event-based estimation (and control), where new measurement samples are not triggered periodically in time but at well-designed critieria on the measurement signal itself or on the current estimation results. In particular, the tutorial discusses key-aspect of event-based system design (triggering criteria, existing algorithms for estimation, distributed architectures) and highlight several successful experimental applications in networked control and robotics. More Details

T2 Relatives for the Kalman filter for tracking and fusion Presenter: Dietrich Fränken Length: 3 hours Brief description: This tutorial gives attendees an overview of Kalman-filter-like estimators for state estimation and fusion while taking a system design point of view. An important aspects of the tutorial are how to approach the options to select between the difference estimators, what will be the resulting effort to put in with respect to both implementation and brain power and what possible side-effects can occur when using those filters too naively. More Details

T3 Robust Kalman filtering Presenter: Benjamin Noack and Florian Pfaff Length: 3 hours Brief description: The tutorial covers various Kalman-filter-like estimators when the observed process has high uncertainties and therefore cannot be captured accurately with a linear models and perfectly selected noise distributions. The tutorial focusses on hybrid estimators relying on the combination of stochastic and set-membership approaches. Several solutions will be presented during this tutorial along with new challenges showing the versatility of hybrid estimation. More Details

T4 Multisensor fusion in Automated driving Presenter: Jos Elfring Length: 3 hours Brief description: This tutorial explains the domain of (cooperative) automated driving from a multisensor fusion perspective. Topics that will be discussed are, among others, Bayesian filtering and track-to-track fusion algorithms in automated driving, domain specific multisensor fusion requirements and prediction and measurement models. Both theoretical considerations and practical constraints for applying multisensor fusion in (cooperative) automated driving will be discussed. More Details

T5 Fusion and Bayesian Reasoning with Subjective Logic Presenter: Audun Jøsang Length: 3 hours Brief description: This tutorial gives attendees an introduction to subjective logic and how it applies to Bayesian network modelling and information fusion. Specific elements of the tutorial are (1) Representation and interpretation of subjective opinions, (2) Algebraic operators of subjective logic, and (3) Applications of subjective logic like Bayesian network modelling and analysis. More Details

T6 Passive Surveillance - Advanced Algorithms and Challenging Applications Presenter: Wolfgang Koch Length: 3 hours Brief description: The tutorial covers partly material of the recently published book of the presenter (Tracking and Data Fusion, Springer 2014) and thus provides a guided introduction to deeper reading with a particular focus on passive surveillance. Starting point is the well-known JDL model of sensor data and information fusion that provides general orientation within the world of fusion methodologies and its various applications, covering a dynamically evolving field of ever increasing relevance. Using the JDL model as a guiding principle, the tutorial introduces into advanced fusion technologies based on practical examples taken from real world applications. More Details

T7 Extended object tracking: Theory and applications Presenter: Karl Granström, Stephan Reuter, and Marcus Baum Length: 3 hours Brief description: The tutorial will introduce the topic of extended object tracking, i.e., object tracking using modern high resolution sensors that give multiple detections per object. State of the art theory will be introduced, and relevant real-world applications will be shown. The theory is presented in several tracking examples with different object types (e.g., pedestrians, bicyclists and cars) and different sensor systems (e.g., Lidar, radar and camera). More Details

T8 Proactive Optimal Control to Infer Information Faster Presenter: Renaldas Urniezius Length: 6 hours Brief description: Assume that we a have a robot and we know what force is applied to the mass center of the robot. Our inputs are the velocity of the robot and the known control force signal. We know that transient signals can tell us about the mass and friction coefficient based on boundary conditions of Newton’s second law. Putting it simpler, the start of the any transient process after switching on a control force tells us what the inertia of the robot is and the end time series provide information about steady state, which explains us, what the friction coefficient is. However, how to incorporate both start and end time series in a single inference step when mass and friction are changing in time? This tutorial answers this question. More Details