This paper is dedicated to solving the problem of concept drift in industrial plants using artificial intelligence methods. For this purpose, methodological approaches and procedures are considered and analyzed. Based on the findings, reference architectures were developed at different abstraction levels that can be used in an industrial environment and enable continuous machine learning. Continuous machine learning offers the possibility of adapting to dynamic changes in the production environment, which are reflected in constantly changing data sets. Through a combination of machine learning techniques, a novel and practical framework for continuous learning, also known as lifelong learning, is presented. The integration of problem-focused machine learning methods is advancing in production, e.g., predictive maintenance, process optimization or fault detection. Thereby, fully or semi-automated adaptations to changing environments requiring continuous improvements are less often explored, although practical use cases often require adaptive capabilities as the physical data distribution may change over time. In this paper, the application was continuously improved based on case studies and empirical results, and finally validated with a quality assurance application. Various methods and approaches for detecting concept and data deviations, retraining, packaging and model updating had to be investigated, which led to the question of what a real industry-oriented implementation could look like. The result is a reference architecture that can run on cloud and edge computing resources. This reference architecture is validated in real-world application in the parquet production sector, proving its feasibility and efficiency.

Summary State‐of‐the‐art science gateways can be connected to several distributed computing infrastructures (DCIs) and are able to run jobs and workflows simultaneously in all those DCIs. Flexibility of accessing diverse data storages from these workflows and assisting end users to manage these storages are however the missing features in current gateway implementations, in which these problems often prove to be a barrier of exploiting the power of distributed computing by user communities having no or little IT competence. This paper addresses these issues by integrating a data bridging service called Data Avenue into WS‐PGRADE/gUSE portal framework. Data Avenue offers tools for the end users to easily manage their data residing on various storage resources, and also, jobs become capable of accessing different storages regardless of the particular distributed computing infrastructure where the job is currently being run. Copyright © 2014 John Wiley & Sons, Ltd.

SUMMARY Maintenance of aging legacy COBOL systems is labor‐intensive and is becoming a more and more difficult problem. Program slicing is a potentially useful analysis for aiding different maintenance activities, including program comprehension, reverse engineering, debugging, and testing. Numerous techniques have been proposed in the past decades; however, in interactive contexts, we found none of them suitable for slicing industrial‐scale COBOL systems due to their large space or preprocessing requirements. This paper proposes a novel static program slicing approach, which is based on context‐sensitive token propagation over control flow graphs (CFGs). CFGs require less space compared with program dependence graphs (PDGs) used by other techniques, and the token propagation method computes the necessary information only, on demand. Algorithms are presented for data flow and full slicing to calculate precise program slices. Preliminary application experiences on industrial‐scale COBOL systems are also summarized. Copyright © 2011 John Wiley & Sons, Ltd.

We introduce an improved software regression testing method that determines the test cases that must be rerun to fulfil a given testing criterion. The key part of our method is regression slicing. A regression slice contains the statements whose state may be changed in the modified program with respect to the original for a given test case.

An integrated testing criterion is proposed that extends traditional criteria to be effective to reveal domain errors. The method requires many fewer test cases and is applicable for any kind of predicates. An automated test data generation algorithm is developed to satisfy the criterion. This is the first integrated algorithm that unites path selection and test data generation. The method is based on function minimization and is extended to find required test cases corresponding to ON-OFF points very quickly. In this way the algorithm is dynamic and thus can be used in practice.