Skip to main content

Automotive Production Engineering Unified Perspective based on Data Mining Methods and Virtual Factory Model

Final Report Summary - AUTOUNIMO (Automotive Production Engineering Unified Perspective based on Data Mining Methods and Virtual Factory Model)

The AutoUniMo project aims the research in the area of application of advanced IT tools and engineering methods in automotive with a special focus on flexible, agile and energy efficient production systems. It joins together researchers from academia and industry that represent different fields of expertise, in order to promote interaction in this interdisciplinary area. The AutoUniMo consortium assembles two European academia partners: Technische Hochschule Ingolstadt (THI) that focus on automotive engineering and Silesian University of Technology (SUT) that focus on advanced IT methods. Two industrial partners share their practical experiences: Continental Ingolstadt is focused on Advanced Driver Assistant Systems (ADAS) and AIUT Gliwice is focused on Automatic Control Systems for Automotive Production.
The scientific domain of the AutoUniMo project is a fusion between emerging research areas in engineering like Component Based Automation, Digital Factories and advanced IT methods like Data Mining, Artificial Intelligence and Multi-Agent Systems. Researches are conducted in the fields of: Data Mining in Manufacturing Execution Systems applied in short series production, Digital Factories and Digital Engineering with focus on fusion between theoretical models and real data collected during production execution, Unified Communication Interfaces applicable for industrial computer systems, Advanced Driver Assistance Systems (ADAS) and last but not least energy efficient production with focus on control systems used by automotive industry.
The main research areas of AutoUniMo project are:
- Effective and flexible Manufacturing Execution Systems (MES) that support short series production in following activities: job scheduling, launching of the orders, response to random events, adjustments of production plans, product genealogy tracing, production quality management and maintenance management activities. The research focus is on the fusion between conceptual and functional models defined by ANSI/ISA-95 (IEC/ISO 62264) norm and advanced IT technologies including Data Mining, Artificial Intelligence and Multi-Agent Systems. The main application areas are: production capacity forecasting, production scheduling, product genealogy tracing and production quality management. The research goals aim the application of data mining and artificial intelligence methods for MES systems dedicated for short series production systems. The software model for Short Series Production MES and related software mockup have been prepared and verified against real production use cases defined by Continental Ingolstadt. One of the practical achievements is the software prototype for MES that support production operations performed by Prototyping Department of ADAS Division of Continental.
- Energy efficient control systems with focus on production stations used by automotive industry. The research aims a fusion between classical control systems based on Programmable Logic Controllers (PLC) with IEC 61131-3 based software and data mining methods used for assessment of energy consumption. The use cases defined by AIUT reflect the experiences from real production stations used by automotive production companies. A special focus was given to multi-variant and short series production systems including automatic production stations and production lines typical for automotive industry. The methodology and supporting software tools has been prepared. The practical achievements of conducted research are hardware components and software tools dedicated for Energy Efficient Control Production Systems in the automotive industry.
- Digital Factories and Digital Engineering with focus on connection between virtual components and real data gathered from automotive production process. The virtual components abstractly describe the functionalities provided by real production process. Research focus is on the new concepts applied in automotive engineering including virtual components used for system planning and virtual commissioning, AML (Automotive Markup Language) application for multi-domain system description, open communication interfaces with context based information models used by service oriented middleware like OPC UA (IEC 62541 standard for OPC Unified Architecture). The digital factory model for automotive production systems has been developed and a set of informatics tools has been created. The digital factory system model for automotive production with functional and nonfunctional requirements analysis has been created. The model describes different phases of preparation of automotive production system including model based decomposition of the system, its description accordingly to Componet Based Technology paradigm. One of the practical achievements is software Framework for Digital Factory applicable in automotive production systems preparation and validation.
- Advanced Driver Assistance Systems (ADAS) with focus on sensor fusion supported by IT methods like Artificial Intelligence (AI) and decision rules systems. The main research focuses on hardware and software components of ADAS, sensor fusion, information processing and communication between ADAS components. The fusion between generic AI and DM approaches and ADAS has been supported by set of functional and nonfunctional requirements analysis of ADAS including its particular components and interactions with other systems. The research case study on the best available DM and AI technologies able to support ADAS systems requirements has been conducted. One of the practical achievements is research environment for AI in ADAS, that has been realized as embedded systems for mobile platform control, evaluation of AI methods for energy efficiency purposes, real time communication of the sensor/actuator values between the mobile platform and CANoe/OPC UA, analysis and implementation of real-time CAN communication between devices as an examples of solution used in Automotive. This environment is used for research and didactic activities by Institute of Informatics, Silesian University of Technology.
- The exchange of research experiences was strengthened by intersectoral secondments. International staff exchange has played a main role (80,1% of total secondment time) while intranational has played a supporting role (19,9% of total secondment time). The total number of performed secondments is equal 151 person/month and includes the participation of 12 researchers: six experienced researchers (ER) and six Early Stage Researchers (ESR). Each host institution has shared knowledge related to its main Research Areas: SUT has provided a knowledge in Data Mining and Unified Communication areas and also has given the support for Artificial Intelligence, Agent Based and Component Based Industrial Computer Systems and IT System Security. THI has supported knowledge related to Virtual Factories and Digital Engineering and also experiences in component based engineering, and industrial control knowledge transfer. CONTI has shared practical experiences in Advanced Driver Assistance Systems. AIUT has shared experiences in Automatic Control Systems for Automotive production systems and Energy Efficient Control for Discreet Production.
- The dissemination activities included: Special Session on Cyber Physical Systems in Automotive Area (CPSiA) on 9th International Conference on Computational Collective Intelligence (ICCCI 2017), Nicosia, Cyprus, September, 2017 - Best Special Session Award published by Lecture Notes in Computer Science, Springer 2017;International Conference on Data Mining and Artificial Intelligence Methods for Automotive Systems, Ingolstadt, Germany, Ingolstadt, 20th April 2017; International Conference on Data Mining and Artificial Intelligence Methods for Automotive Systems, May 23-24, 2016, Szczyrk, Poland; International Seminar on Data Mining and Artificial Intelligence Methods for Automotive Systems, November 26th, 2015, Silesian University of Technology, Gliwice, Poland;International Seminar on Emerging Technologies in Automotive Area, Ingolstadt, Bavaria, Germany, 4th December 2014; AutoUniMo Scientists Night 2017, Ingolstadt, Germany; AutoUniMo Open Day 2017, Ingolstadt, Germany; AutoUniMo Scientists Night 2016, Gliwice, Poland; AutoUniMo Open Day 2016, Gliwice, Poland
Altogether, the research program resulted in: 30 scientific publications, two scientific conferences, two seminars and four outreach activities. The full list of publications and other dissemination activities is available on the project webpage with most of them being available for download.
Project website: