Research Field C: Lokalisation von Mikrosystemen

The ability of self-localization is a prerequisite for many microsystem applications. The output of this research field is sensors, methods and algorithms.

Projekte im Forschungsbereich C Localization of Microsystems:

C.1

Integrated Magnetic Sensor Systems for Localization

C.2

Sensor Data Fusion for Embedded Microsystems

C.3

Miniaturized, Robust and Highly Precise Local Positioning Systems

C.1 Integrated Magnetic Sensor Systems for Localization

The aim of this project is to develop highly sensitive magnetic sensor systems taking advantage of the Hall effect or magnetoelectromechanical effects in CMOS structures. Hall-effect-based high-sensitivity systems require the deep understanding of offset effects, optimized design methods, and the implementation of compensation techniques with adequate integrated circuits. The magnetic field sensor systems resulting from this project will provide some of the sensory hardware for the research topics of this PhD program in the field of localization. The reason for trying to achieve the highest possible sensitivities is that this makes possible the use of less dense sensor networks for localizing magnetic objects. A possible further spin-off is the realization of an integrated compass.

Contact:
Prof. Dr. Oliver Paul
Chair of Microsystem Materials
Department of Microsystems Engineering
Georges-Köhler-Allee 103, 79110 Freiburg
Phone: +49-761-203-7191 | Fax: +49-761-203-7192 | E-mail

C.2 Sensor Data Fusion for Embedded Microsystems

If embedded microsystems are equipped with sensors, e.g. sensor arrays or positioning sensors, one important question is, how to interpret the data provided by the sensors. Because microsystems have only limited communication capabilities, the data have to be filtered and also reduced on the system before they can be transmitted. In this context it is very important to apply filter algorithms that only demand low computation resources. In this project we will consider approximations of standard filtering algorithms that allow the implementation under the given constraints of an embedded microsystem. At the same time we will also consider how to model the dependency between measurements of different sensors. This, for example, plays a crucial role in the context of sensor arrays. By appropriately modelling the dependencies of the sensors we will be able to obtain better results compared to a system that considers all measurements as independent. To achieve this, we will introduce probabilistic sensor models for the sensor systems in embedded microsystems. We then will develop efficient filter algorithms that take the dependencies of the sensors into account. These problems will be studied in the context of positioning sensors as well as sensor arrays.

Contact:
Prof. Dr. Wolfram Burgard
Chair of Autonomous Intelligent Systems
Department of Computer Science
Georges-Köhler-Allee 079, 79110 Freiburg
Phone: +49-761-203-8026 | Fax: +49-761-203-8007 | E-mail

C.3 Miniaturized, Robust and Highly Precise Local Positioning Systems

The growing integration of analog high-frequency components in highest integrated digital circuits, rising from the rapid development of mobile radio systems, enable further wireless services in the industrial, private and medical environment. It is expectedly that the boom of mobile communications is followed by a boom of radar technology that can change our society in a way similar the mobile phones did.

The local positioning system (LPS) that is developed in this subproject allows the contactless real-time measurement of the coordinates of microsystem positioning sensors (backscatters, transponders). In contrast to GPS systems, the target usage of this system is in the range of 10 to 100 meters like inside of a hall or on a football place. But, to allow any usages in control systems it is necessary that the resolution of the LPS system is in the order of 1 to 10 centimetres. Additionally, the system has to be robust concerning electromagnetic interference and fading.

To reach these targets, a combination of absolute localization techniques must be combined with relative sensor measurements. The absolute localization can be implemented by using fixed RF-ID tags or using time of flight measurements with the help of a micro radar and active backscatters. For special applications an optical system is also suited. All these systems, however, can not ensure a permanent localization since they all can be blocked by shadowing.

In this case relative sensors consisting of gyros, accelerometers, or inertial sensors might be used to track forward the position. The coupling between the absolute localization and the relative sensors may be implemented by, e.g., using an extended Kalman filter.

The target of this subproject is

• to evaluate the specification of each single component used in a LPS
• to specify, design, and test a LPS using microsystem absolute and relative sensors
• to export the positioning data to a navigation software
• to probe the capability of the system on special application fields.

Contact:
Prof. Dr. Leonhard Reindl
Chair of Electrical Instrumentation
Department of Microsystems Engineering
Georges-Köhler-Allee 103, 79110 Freiburg
Phone: +49-761-203-7220 | Fax: +49-761-203-7222 | E-mail