Scenarios

Embedded systems are hardware and software systems that are integrated in ambient technical systems and that perform complex control and data processing tasks. Typical application areas of embedded systems are automobiles, aerospace, telecommunication, monitoring and control of manufacturing processes and instrumented homes. The miniaturization of the embedded systems and the addition of sensory capabilities, i.e., the development of embedded microsystems, are envisioned to enable an even broader range of application to be addressed. The smaller the individual components are the more flexibility and mobility can be achieved. Furthermore the cost of embedded systems will be reduced and thus facilitates the enhancement of more and more objects with processing power and sensing capabilities. Further innovative applications can be developed when embedded microsystems possess the ability to communicate among each other and with external computers. Several important questions in the context of embedded microsystems become apparent when one considers the following three scenarios.

go directly to: Scenario 1 - Scenario 2 - Scenario 3 - Discussion

Scenario 1

The automation of production lines typically requires serious capital investments. Especially shorter product life cycles and the increasing diversity of product variants require flexible production facilities. The usually central information processing leads to extensive data streams. For example, when bar codes are used to identify products, one has to generate an appropriate query to a central data base to retrieve the relevant product information. By means of embedded microsystems, however, this information can be stored directly with the product. The manufacturing engine can then directly retrieve the information from the product and can adapt itself to the specific requirements. The self localization of a product furthermore allows a decentralized organization of the material flow including the of product storage.

Scenario 2

A further example is a system of embedded microsystems attached to the human body. Such a system can permanently carry out diagnostic tasks and even initiate therapeutical action. In contrast to already existing systems such as devices for monitoring blood pressure, embedded microsystems allow a direct and continuous monitoring. If necessary, they can even communicate with other diagnostic tools in the home of the person or when visiting a doctor.

Scenario 3

Multimedia techniques play an important role in the context of teaching. Many modern systems for recording lectures are based on dedicated hardware, such as a touch screen or an electronic whiteboard. The integration of traditional techniques such as chalk and blackboard is achieved only at the cost of a significant effort using sophisticated image processing capabilities or dedicated and expensive sensors. With a microsystem that is embedded in the chalk and allows a reliable localization, however, one can directly record the text written by the teacher to the blackboard without any modifications of the environment.

Discussion

Although these scenarios represent only examples, they still demonstrate the utility and the enormous application relevance of embedded microsystems. Such systems will pervade many areas of our life and will not only offer new functionalities but will inconspicuously and reliably fulfill their tasks. The application of embedded microsystems in medicine and in the consumer market will increase the quality of life; the utilization in technical systems of automation or automotive engineering will lead to higher flexibility with reduced investment costs. In this way they will contribute to the competitiveness of the Federal Republic of Germany.

To reach this ambitious goal we need to solve important research questions, which require the interdisciplinary collaboration between microsystem technology and computer science. The Faculty for Applied Sciences of the University of Freiburg with its two Institutes for Microsystem Technology and Computer Science under one common roof constitutes an ideal institution for such an effort. Since the interdisciplinary area of embedded microsystems is not covered by any existing education program, the University of Freiburg appears to be an ideal site for a post graduate program in this field.

Whereas all three scenarios mentioned above illustrate a high application potential of embedded microsystems, they also highlight important research topics. The first scenario, for example, illustrates the relevance of the localization problem, which amounts to estimating the pose of objects in space. Depending on the particular domain, certain modalities or combinations of them will be more appropriate (e.g., microwave time of flight measurement and estimation of magnetic fields). This, however, requires appropriate techniques for the data fusion and object tracking. Within the PhD program Embedded Microsystems, the localization of embedded microsystems plays a key role. The research topics range from the development of sensitive magnetic field sensors to efficient techniques for data fusion and requires the tight cooperation between microsystem technology and computer science.

In the context of the medical diagnosis it is typically assumed that multiple embedded microsystems distributed over the space, have to fulfill a joint task. Example scenarios are the organization of a supply chain or the diagnosis of cardiovascular functions of a patient. Applications of this kind require key technologies that are covered in the research area Distributed Microsystems: intelligent microsystems, communication in changing configurations, and fulfillment of tasks under energy constraints. In interdisciplinary projects we plan to develop and analyze microsystems communicating over wireless networks as well as efficient communication architectures. In the view of the application scenarios we plan to develop distributed microsystems for instrumentation applications and for energy efficient automation.

The increasing complexity of embedded microsystems also requires efficient design methods. For example, when manufacturing sequences need to be changed, the software must be quickly adaptable. Potential tools are the automatic generation of operating systems tailored to the application and the architecture. Additionally, one requires effective tools for the development of hardware. Implanted microsystems for diagnostic purposes, for example, require specific, energy efficient microcontrollers. In this context, the integrated consideration of the complex interplay of analog and digital technology, of sensors and data processing constitutes an important challenge. In the research area Modelling and Development of this program we will develop specific development tools spanning across all levels of hardware and software development.

Especially the application Scenario 2 (medical technology) illustrates, that security is a central aspect of embedded microsystems. With the increasing complexity of the systems, however, it becomes even harder to guarantee their security and safety. Conventional techniques for the verification of digital circuits are of limited applicability to embedded microsystems, since such systems may contain analog and mechanical components. The consortium therefore puts a major emphasis on the area of Diagnosis and Test of embedded microsystem. For the development process of microsystems, it is planned to investigate test procedures for electronic and mechanical components. Furthermore, reliable microsystems require the ability to monitor their own status and perform self-diagnosis during their operation.