Preliminary Technical Program




Sessions


Session 1. Design methods and Model-Driven Engineering (105 min)
1. Communication Layer architecture for a production line Digital Twin using Hierarchical Colored Petri Nets (short paper)
2. Evaluating the Perceptual Properties of Crosscutting Concerns Occurrence Points Specifications in Embedded Software
3. Simulation of Timing Attacks and Challenges for Early Side-Channel Security Analysis
4. Adaptation for Energy Saving in Time-triggered Systems Using Meta-scheduling with Sample Points

Session 2. Hardware architectures for embedded systems I (75 min)
1. Exploiting heterogeneity in PIM architectures for data-intensive applications
2. Demonstrating Scalability of the Checkerboard GPC with SystemC TLM-2.0
3. MAFAT: Memory-Aware Fusing and Tiling of Neural Networks for Accelerated Edge Inference (short paper)

Session 3. Engineering Applications of Artificial Intelligence targeting embedded systems (90 min.)
1. Analysing the Characteristics of Neural Networks for the Recognition of Sugar Beets
2. Synthetic Data for Machine Learning on Embedded Systems in Precision Agriculture
3. Using Network Architecture Search for Optimizing Tensor Compression

Session 4. Hardware architectures for embedded systems II (60 min.)
1. Memristor-only LSTM acceleration with non-linear activation functions
2. Minimizing Memory Contention in an APNG Encoder using a Grid of Processing Cells



Keynotes


(03.11.2022)
Safe Integration of Learning in Embedded Systems
Prof. Dr. Paula Herber

Embedded systems are ubiquitous in our daily lives, and they are often used in safety-critical applications, such as cars, airplanes, or medical systems. As a consequence, there is a high demand for formal methods to ensure their safety under all circumstances. A prerequisite for the application of formal methods, however, is a formal model of the system under verification. Such a formal model is typically not available in industrial design processes, where informally defined modeling and design languages such as SystemC or Simulink are used. Furthermore, systems that learn their behavior in trial-and-error processes - which is gaining enormous importance in increasingly autonomous embedded systems - are extremely hard to capture formally. In this talk, I will summarize our work on automated transformations from informal system descriptions into formal languages and on the use of contracts to safely integrate learning components in the formal verification process.



(04.11.2022)
How Connectivity contributes to Smart Farming
Prof. Dr. Andreas W├╝bbeke

Connectivity for agricultural machines is not a brand new topic, but with every new network generation the possibilities grow to make life easier and smarter for farmers. In this talk we want to provide an overview on possibilities and already started actions to improve farming by using mobile networks. Thereby collecting data of agricultural machinery for optimizing individual machines as well as machine fleets is in focus. Today the transformation of this machine data to models by the help of big data analysis and machine learning is the key competence for smart farming.



Towards Models for Smart Cyber-Physical Systems of Systems
Prof. Dr. Holger Giese

Currently, a dramatic transformation of our technical world towards smart cyber-physical system of systems can be observed. Networked embedded systems with their interaction with the physical world are enhanced to system of systems interconnected in the cyber. Furthermore, employing artificial intelligence techniques and in particular machine learning enables smart system of systems. This raises the question whether our models for these future embedded systems are ready to tackle the resulting challenges. Models are used in classical engineering to plan a single system. For smart cyber-physical system of systems at first the scope changes from a single system to a system of systems and its systems. Furthermore, in classical engineering the models are used upfront to maximize envisioned properties resp. minimize cost, while for smart cyber-physical system of systems these development-time considerations are not enough and in addition also models at run-time as a means for context- and self-awareness are employed and learned. In this keynote, we will first review the underlying causes for these shifts and outline related open challenges and implications for future models to engineer smart cyber-physical system of systems. Then, we present our work on models for designing, analyzing, and learning for the envisioned future cyber-physical system of systems. Finally, we will discuss to what extent our current capabilities match the beforehand identified challenges and where substantial improvements will be crucial in the future.