SS01 Big Data and Semantic Technologies for Automation
Special Session Organizers:
Sören Volgmann, Fraunhofer Application Center Industrial Automation (IOSB-INA)
Alexander Maier, Fraunhofer Application Center Industrial Automation (IOSB-INA)
Prof. Dr. Oliver Niggemann, OWL University of Applied Sciences
The special session comprises two sub-topics, Big Data technologies as well as Semantic Technologies for Automation:
Big Data Technologies for Automation
Aim: In the future, the Internet of Things (IoT) will deliver a growing amount of data from production plants, rendering a manual analysis of these data impossible Therefore, any implementation of the Internet of Things will require corresponding methods for data analysis and machine learning. I.e. what is needed are Big Data solutions for the field of automation. But the IoT demands, for several reasons, other solutions than classical big data scenarios: E.g. measurements are time-dependent and unsynchronized. Furthermore, all algorithms must be suited for embedded platforms.
Topics: We invite submissions of papers addressing aspects of data analysis, machine learning and Big Data in the field of automation and production. This part of the special session will focus on (but not be limited to) the following topics in the area of:
- Data acquisition strategies in distributed plants
- Methods for storing data in the field of industrial automation
- Machine learning for condition monitoring and diagnosis
- Industrial applications of Big Data and machine learning
- Learning of behaviour models for complex industrial processes
- Unsupervised and supervised learning from large amounts of data
- Visualization of plant behaviour
- Self-optimization of complex industrial processes
Semantic Technologies for Automation
Aim: Industrial production is extending its spectrum to include stronger customization of products, both in discrete manufacturing and in process automa-tion. In the corresponding business model, standard factors such as high volumes or maximum efficiency no longer exist, and time to market (TTM) is a highly sensitive factor affecting profitability. Efficient engineering techniques which enable rapid creation from high level specifications to implementation or enable flexible production can be crucial since they greatly shorten the development time.
The goal of this part of the special session is to invite recent research efforts in realizing the dream of “Automation of Automation”, i.e., automating the engineering process for automation systems, in both factory and process automation. Moving towards this direction, one requires close collaboration from researchers in industrial automation (for appropriate semantic models) and in computer science (for efficient algorithms, e.g., graph-database querying, game solving, constraint solving), such that one can create tools that both ease modelling and scale to the industrial usage.
The proposed part of the special session will invite interdisciplinary papers presenting their results in successfully introducing semantic technologies and smart algorithms to reduce manual engineering efforts.
Topics: This part of the special session will be focusing on (but not limited to) the following topics:
- Achieve flexible production or avoid re-engineering via using a well-defined semantic (e.g., behavioral-level) interface and via the introduc-tion of smart searching algorithms to automatically synthesize artifacts (e.g., optimized production plans, parameters, PLC function blocks) that originally needs to be created manually.
- Apply generic automatic reasoning technologies on the semantic model on the device level (bus, PLC, sensors and actuators) to achieve smart manufacturing and distributed intelligence.
- Experience papers stating how semantic technologies and algorithmic search/synthesis/reasoning techniques are used to avoid engineering efforts.
- Position papers from industry prioritizing problems to be solved.
A selection of papers will be invited to a special issue of Semantic Web Journal.