Analyzing Requirements on Software Tools According to the Functional Engineering Phase in the Technical Systems Engineering Process Nicole Schmidt, Arndt Lüder Otto-v.-Guericke University, Universitäts- platz2, 39106 Magdeburg, Germany [nicole.schmidt, arndt.lueder]@ovgu.de Heinrich Steininger logi.cals Austria GmbH, Europaplatz 7/1, 3100 St. Pölten, Austria heinrich.steininger@logicals.com Stefan Biffl Vienna University of Technology Christian Doppler Laboratory CDL-Flex Institute of Software Technology Favoritenstr. 9/188, Vienna, Austria Stefan.Biffl@tuwien.ac.at Abstract The capability to shorten the process of engineering technical systems, from the mission to build the system to the final commissioning of the system, has become a crit- ical factor for company success. This engineering process has become a time and money consuming part of the overall life cycle of a technical system. Common ap- proaches for tackling this problem are the improvement of functions of isolated tools and the improve- ment/provision of the interoperability between tools. However, there is only little information on needs and preferences of domain experts. In this paper we report on a questionnaire-based sur- vey to elicit the needs and preferences of domain experts. We analyze the answers of 88 domain experts from dif- ferent disciplines with a variety of backgrounds on the importance of tool functionalities during functional en- gineering and discuss improvement aspects of three ma- jor functionalities rated as most important: version management, data exchange, and consistency checking. 1. Motivation The combination of demand for reducing the time to market and the continuously increasing complexity of production systems is a new challenge for engineers dur- ing the engineering process of technical systems [1, 28]. But planning is a complex process: Where to start to change? In 2005, a study had analyzed the engineering process from the first action, the mission engineering for a specific production system, to the last action, the commissioning of the system [2]. The result was that about 50% of the overall costs come from the engineer- ing of the production system. But how to reduce the en- gineering effort? Two approaches based on better tool support can be found [3]: Improving the engineering tools: The usability of the engineering tools will be optimized towards the use within a special set of engineering activities reducing the necessary manual engineering efforts. Improving the interoperability of tools within tool chains: The integration of engineering activities crossing tool borders can be improved, consequently reducing the manual effort for data exchange and data representation in data-receiving engineering activities [11]. Interopera- bility of tools can be provided by a tool suite, which con- tains all needed tools aligned to each other, or by setting up a tailored tool chain with the best-fitting but often he- terogeneous tools that need to be integrated effectively and efficiently [4]. In this context the question arises which approach has a higher potential for domain experts to save time and money in an engineering organization, i.e. which ap- proach is worth pursuing and spending money on realiz- ing it? Is it at all possible to make a clear decision? What are relevant factors to support making the decision? To help answering these questions the Vienna Uni- versity of Technology, the software provider logi.cals GmbH, and the Otto-von-Guericke University Magde- burg, all with a research and development scope mainly on control engineering, conducted a survey with domain experts. Engineers from a wide range of disciplines were asked for their opinion about the importance of capabili- ties of engineering tools according to existing and up- coming tool trends relevant for the functional engineer- ing phase. Main goal was getting a list of functionalities and concepts, which the users of engineering tools de- sire. According to this list the authors wanted to find an answer to the initial question, which tool improvement approach is likely to generate a higher benefit. Following the analysis results we consider how these functionalities could be implemented with a modern in- tegration framework, such as the Automation Service Bus [11].