Time filter

Source Type

Weber B.,University of Innsbruck | Mutschler B.,Ravensburg-Weingarten University of Applied Sciences | Reichert M.,University of Ulm
Science of Computer Programming | Year: 2010

Business Process Management (BPM) technology has become an important instrument for supporting complex coordination scenarios and for improving business process performance. When considering its use, however, enterprises typically have to rely on vendor promises or qualitative reports. What is still missing and what is demanded by IT decision makers are quantitative evaluations based on empirical and experimental research. This paper picks up this demand and illustrates how experimental research can be applied to technologies enabling enterprises to coordinate their business processes and to associate them with related artifacts and resources. The conducted experiment compares the effort for implementing and maintaining a sample business process either based on standard workflow technology or on a case handling system. We motivate and describe the experimental design, discuss threats for the validity of our experimental results (as well as risk mitigations), and present the results of our experiment. In general, more experimental research is needed in order to obtain valid data on the various aspects and effects of BPM technology and BPM tools. © 2009 Elsevier B.V. All rights reserved.

Rager M.,Ravensburg-Weingarten University of Applied Sciences | Gahm C.,University of Augsburg | Denz F.,University of Augsburg
Computers and Operations Research | Year: 2015

Energy efficiency has become more and more critical for the success of manufacturing companies because of rising energy prices and increasing public perception of environmentally conscious operations. One way to increase energy efficiency in production is to explicitly consider energy consumption during short-term production planning. In many cases, final energy sources (FES) are not directly consumed by production resources and thus have to be transformed by conversion units into applied energy sources (AES), such as steam or pressure, so the relationship between AES and FES has to be considered. Therefore, we present an energy-oriented scheduling approach for a parallel machine environment. These parallel machines require production order and process time specific amounts for AES and the objective is to minimize the demand of FES. This minimization can be achieved by smoothing the cumulated demand of AES to avoid the frequent load alternations that are responsible for the inefficient operation of conversion units. Therefore, resource leveling is used as a surrogate objective for optimization. To solve the resource leveling problem for large problems, a Genetic Algorithm and two Memetic Algorithms are developed. The evaluation of the proposed Evolutionary Algorithms is based on small test instances and several real-world instances. These latter instances are based on an application case from the textile industry, and promising results concerning energy costs and carbon dioxide emissions are reported. © 2014 Elsevier Ltd. All rights reserved.

Tokic M.,University of Ulm | Tokic M.,Ravensburg-Weingarten University of Applied Sciences | Palm G.,University of Ulm
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2011

This paper proposes "Value-Difference Based Exploration combined with Softmax action selection" (VDBE-Softmax) as an adaptive exploration/exploitation policy for temporal-difference learning. The advantage of the proposed approach is that exploration actions are only selected in situations when the knowledge about the environment is uncertain, which is indicated by fluctuating values during learning. The method is evaluated in experiments having deterministic rewards and a mixture of both deterministic and stochastic rewards. The results show that a VDBE-Softmax policy can outperform ε-greedy, Softmax and VDBE policies in combination with on- and off-policy learning algorithms such as Q-learning and Sarsa. Furthermore, it is also shown that VDBE-Softmax is more reliable in case of value-function oscillations. © 2011 Springer-Verlag.

Mutschler B.,Ravensburg-Weingarten University of Applied Sciences | Reichert M.,University of Ulm
Studies in Computational Intelligence | Year: 2013

Providing effective IT support for business processes has become crucial for enterprises to stay competitive in their market. Business processes must be defined, configured, implemented, enacted, monitored and continuously adapted to changing situations. Process life cycle support and continuous process improvement have therefore become critical success factors in enterprise computing. In response to this need, a variety of process support paradigms, process specification standards, process management tools, and supporting methods have emerged. Summarized under the term Business Process Management (BPM), they have become a successcritical instrument for improving overall business performance. However, introducing BPM approaches in enterprises is associated with significant costs. Though existing economic-driven IT evaluation and software cost estimation approaches have received considerable attention during the last decades, it is difficult to apply them to BPM projects. In particular, they are unable to take into account the dynamic evolution of BPM projects caused by the numerous technological, organizational and project-specific factors influencing them. The latter, in turn, often lead to complex and unexpected cost effects in BPM projects making even rough cost estimations a challenge. What is needed is a comprehensive approach enabling BPM professionals to systematically investigate the costs of BPM projects. This chapter takes a look at both known and often unknown cost factors in BPM projects, shortly discusses existing IT evaluation and software cost estimation approaches with respect to their suitability for BPM projects, and finally introduces the Eco- POST framework. EcoPOST utilizes evaluation models to describe the interplay of technological, organizational, and project-specific BPM cost factors as well as simulation concepts to unfold the dynamic behavior and costs of BPM projects. © Springer-Verlag Berlin Heidelberg 2013.

Michelberger B.,Ravensburg-Weingarten University of Applied Sciences | Mutschler B.,Ravensburg-Weingarten University of Applied Sciences | Reichert M.,University of Ulm
Proceedings of the 2012 IEEE 16th International Enterprise Distributed Object Computing Conference, EDOC 2012 | Year: 2012

Today, enterprises are confronted with a continuously increasing amount of data. Examples of such data include office files, e-mails, process descriptions, and data from process-aware information systems. This data overload makes it difficult for knowledge-workers to identify the information they need to perform their tasks in the best possible way. Particularly challenging is the alignment of process-related information with business processes. In fact, process-related information and business processes are usually managed separately. On the one hand, enterprise content management systems, shared drives, and Intranet portals are used for organizing information, on the other hand, process management technology is used to design and enact business processes. With process-oriented information logistics (POIL) this paper presents an approach for bridging this gap. POIL enables the process-oriented and context-aware delivery of process-related information to knowledge-workers. We also present a clinical use case and a proof-of-concept prototype to demonstrate the application and benefits of POIL. © 2012 IEEE.

Tokic M.,Ravensburg-Weingarten University of Applied Sciences | Tokic M.,University of Ulm
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2010

This paper presents "Value-Difference Based Exploration" (VDBE), a method for balancing the exploration/exploitation dilemma inherent to reinforcement learning. The proposed method adapts the exploration parameter of ε-greedy in dependence of the temporal-difference error observed from value-function backups, which is considered as a measure of the agent's uncertainty about the environment. VDBE is evaluated on a multi-armed bandit task, which allows for insight into the behavior of the method. Preliminary results indicate that VDBE seems to be more parameter robust than commonly used ad hoc approaches such as ε-greedy or softmax. © 2010 Springer-Verlag Berlin Heidelberg.

Kurniawan T.A.,Ravensburg-Weingarten University of Applied Sciences | Kurniawan T.A.,University of Eastern Finland | Sillanpaa M.E.T.,University of Eastern Finland | Sillanpaa M.,Finnish Environment Institute
Critical Reviews in Environmental Science and Technology | Year: 2012

The authors present an overview with critical analysis of technical applicability of various nanoadsorbents such as carbon nanotubes, nano-zerovalent iron, and metal oxides-based and polymeric nanoparticles in treating contaminated water. To highlight their performance, selected information such as synthesis method, pH, dose required, pollutant's concentrations, reaction time, and treatment efficiency is presented based on the literature survey of 276 articles (1989-2010). Their advantages and drawbacks in applications are evaluated. Nanoadsorbents that stand out for outstanding performance are compared to bulk activated carbon. The implications of nanoadsorbents to public health and their way forward for facilitating environmental sustainability are also discussed. © 2012 Taylor & Francis Group, LLC.

Schneider M.,Ravensburg-Weingarten University of Applied Sciences | Ertel W.,Ravensburg-Weingarten University of Applied Sciences
IEEE/RSJ 2010 International Conference on Intelligent Robots and Systems, IROS 2010 - Conference Proceedings | Year: 2010

In recent years there was a tremendous progress in robotic systems, and however also increased expectations: A robot should be easy to program and reliable in task execution. Learning from Demonstration (LfD) offers a very promising alternative to classical engineering approaches. LfD is a very natural way for humans to interact with robots and will be an essential part of future service robots. In this work we first review heteroscedastic Gaussian processes and show how these can be used to encode a task. We then introduce a new Gaussian process regression model that clusters the input space into smaller subsets similar to the work in [11]. In the next step we show how these approaches fit into the Learning by Demonstration framework of [2], [3]. At the end we present an experiment on a real robot arm that shows how all these approaches interact. ©2010 IEEE.

Vien N.A.,Ravensburg-Weingarten University of Applied Sciences | Ertel W.,Ravensburg-Weingarten University of Applied Sciences | Chung T.C.,Kyung Hee University
Applied Intelligence | Year: 2013

This paper considers the problem of extending Training an Agent Manually via Evaluative Reinforcement (TAMER) in continuous state and action spaces. Investigative research using the TAMER framework enables a non-technical human to train an agent through a natural form of human feedback (negative or positive). The advantages of TAMER have been shown on tasks of training agents by only human feedback or combining human feedback with environment rewards. However, these methods are originally designed for discrete state-action, or continuous state-discrete action problems. This paper proposes an extension of TAMER to allow both continuous states and actions, called ACTAMER. The new framework utilizes any general function approximation of a human trainer's feedback signal. Moreover, a combined capability of ACTAMER and reinforcement learning is also investigated and evaluated. The combination of human feedback and reinforcement learning is studied in both settings: sequential and simultaneous. Our experimental results demonstrate the proposed method successfully allowing a human to train an agent in two continuous state-action domains: Mountain Car and Cart-pole (balancing). © 2013 Springer Science+Business Media New York.

Voos H.,Ravensburg-Weingarten University of Applied Sciences | Bou-Ammar H.,Ravensburg-Weingarten University of Applied Sciences
Proceedings of the IEEE International Conference on Control Applications | Year: 2010

Quadrotor UAVs are one of the most preferred type of small unmanned aerial vehicles because of the very simple mechanical construction and propulsion principle. However, the nonlinear dynamic behavior requires a more advanced stabilizing control and guidance of these vehicles. In addition, the small payload reduces the amount of batteries that can be carried and thus also limits the operating range of the UAV. One possible solution for a range extension is the application of a base station for recharging purpose even during operation. In order to increase the efficiency of the overall system further, a mobile base station will be applied here. However, landing on a moving base station requires autonomous tracking and landing control of the UAV. In this paper, a novel nonlinear autopilot for quadrotor UAVs is extended with a tracking and landing controller to fulfill the required task. First simulation and experimental results underline the performance of this new control approach for the current realization. © 2010 IEEE.

Loading Ravensburg-Weingarten University of Applied Sciences collaborators
Loading Ravensburg-Weingarten University of Applied Sciences collaborators