188宝金博页面版

Custom-Developed vs. Model-based Configuration Tools: Experiences from an Industrial Automation Ecosystem Daniela Lettner 1 Michael Petruzelka 2 Rick Rabiser 1 Florian Angerer 1 Herbert Prähofer 1 Paul Grünbacher 3 1 CDL MEVSS Johannes Kepler University Linz 4040 Linz, Austria daniela.lettner@jku.at 2 KEBA AG Gewerbepark Urfahr 4041 Linz, Austria pet@keba.com 3 Systems Engineering and Automation Johannes Kepler University Linz 4040 Linz, Austria paul.gruenbacher@jku.at ABSTRACT High demands regarding the variability of automation software motivate organizations to automate the configuration process. In practice, this often leads to the development of custom configura-tion tools designed specifically for configuring the automation software they were developed for. This approach works well as long as both, the development of the software and the configura-tor are under the full control of the organization. However, soft-ware platforms are increasingly open, i.e., key customers add capabilities and thereby change the platform’s variability. Often, these customers create a new platform themselves, which they offer to their customers. Moving from a closed platform to a soft-ware ecosystem means that development and variability manage-ment happen at multiple layers involving multiple teams with different backgrounds. This poses new requirements regarding the flexibility of configuration tools. In this paper, we report experi-ences and issues with a custom-developed configurator currently in use in an industrial automation software ecosystem. We de-scribe how a model-based tool can be applied to address these issues and provide a scenario-based comparison of the custom-developed solution and the model-based configurator. Categories and Subject Descriptors D.2.13 [Reusable Software]: Reuse Models General Terms Management, Documentation, Design. Keywords Software ecosystem, configuration, model-based product lines. 1. INTRODUCTION Many automation software systems are developed as product lines due to increasing demands regarding their variability. Organiza-tions developing such systems have developed a wide range of techniques to deal with the existing variability. However, compa-nies typically do not apply product line engineering (PLE) [7] methods in a strict form to manage the variability of their soft-ware. As confirmed in a recent survey [4] on variability modeling in industrial practice, organizations start using variability models and tools [8], but "no golden standard is set yet" and "industry is still experimenting with various solutions." For instance, custom-developed solutions designed for a particular application domain are very common to support variability management and product configuration. This approach works well as long as both the de-velopment of the software and the configuration tool are under the full control of the organization. However, an increasing number of automation software product lines today are developed as software ecosystems [5], i.e., the platforms become more open and external as well as internal de-velopment teams are responsible for variability management. In such a development approach customers consume and produce software to create new products. This means that customers need support for customizing product variants as well as support for defining and modeling variability in an easy and straightforward manner. As a result, organizations face challenges regarding the maintenance of their custom-developed configuration tools. Hence, model-based approaches supporting variability manage-ment become more attractive. For instance, KEBA AG 1 has been developing and maintaining a custom-developed configuration tool in their software product line for industrial automation. The tool has successfully been used over several years. Currently, however, their automation software product line is evolving into a software ecosystem. External and internal teams with different backgrounds and knowledge are involved in the variability management and configuration of the system platform, the domain solution, as well as customer-specific applications. This means that the maintenance of the KEBA configuration solution is challenged by numerous new requirements, systems, and people that need to be supported. In the collaboration of KEBA and the MEVSS research lab, we are thus investigating the application of model-based PLE ap-proaches [6][8] in KEBA’s software ecosystem. This paper sum-marizes our experiences to date and discusses the use of custom-developed vs. model-based configuration tools in ecosystems. More specifically, we carefully analyze the existing, custom-developed configuration tool from the perspective of software ecosystems challenges (Section 2) and investigate the steps neces- 1 www.keba.com Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distrib-uted for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Permissions@acm.org. SPLC 2013 workshops, August 26 - 30 2013, Tokyo, Japan Copyright 2013 ACM 978-1-4503-2325-3/13/08…$15.00. 52