Johan Eker

Johan Eker is a member of the research group at Ericsson Mobile Platforms in Lund, Sweden, with focus on embedded real-time software for consumer electronics. He received his Masters in Engineering Physics in 1994 and his PhD in Automatic Control 1999, both at Lund University. The thesis dealt with real-time and control system co-design. At connectBlue he worked with wireless networked control system over Bluetooth in 2000, and after which he joined the Ptolemy group at University of California at Berkeley in 2001, where he looked
at system level components and programming languages. Recent research involves computational models for real-time tasks. He has been with Ericsson Research since 2003.

Embedded Software for Mobile Terminals

Abstract:

Mobile terminals today are getting more complex by the hour. Features are added at an unprecedented speed and the code base is constantly increasing.

Embedded software for consumer electronics is getting increasingly complex, with a mix of real-time and desktop like features. The definition of embedded systems is definitely blurred with the arrival the next generation cell phones, where a 3G terminal consists of several million lines of code supporting use-cases with a large number of parallel activities. With advanced user interfaces and applications, on the surface, they very remind of desktop systems. However, looking under the hood, it is very much about traditional embedded programming, where CPU cycles, memory, and power are scarce resources, and hand coded assembler is necessary to get the desired performance. This is a time consuming and fragile process.

This new, extended embedded space puts high strains on the currently available development tool chains and methodologies. The possibilities for embedded systems to evolve and become more reliable, while yet more complex, to some extent depend on what the next generation real-time operating systems and implementation tools have to offer. The embedded world is in many respects stuck with technology originating in the early 70s, e.g. fixed priority scheduling and the C programming language. New tools and ideas are needed to increase productivity, but most importantly, performance.

Jan Lindblad

Jan Lindblad is System Architect at Enea Embedded Technology, working as the System Manager for the OSE family of embedded operating systems. Previous experiences include Technical Product Marketing and Technical Sales. Before joining Enea, Jan engineered fault tolerant software platforms at Ericsson Software Technology.

Overwhelming complexity

Abstract:

"The number of embedded systems around us is increasing rapidly, and more and more sensitive applications are entrusted embedded systems. The cost of failure is increasing. On the other hand, the system complexity increases as devices get integrated and start to communicate. The analyzability is going down, the cost of testing sky rocketing. Movement in several directions will be necessary to counter these developments:

• We need to raise the abstraction level for systems programmers
• We need methods to analyze and test large systems
• We need the systems to handle unforeseen situations autonomously

Mercè Griera I Fisa

Mercè Griera is project officer at the European Commission's Information Society and Media Directorate General. She works in the Six Framework Programme for Research and Development, concretelly in the area of Embedded Systems. She entered the Commission in 1993 in the area of Research Networking, she moved to High Performance Computing and Networking and to Communications Technologies. Before joining the European Commission, she was technical director of CAETI (Generalitat de Catalunya) and lecturer and researcher at the Autonomous University of Barcelona. Details on how to contact this speaker can be found here.

Real-Time and Embedded Systems: Research Opportunities  in the IST Thematic Priority of the 6FP