ScienceDirect IFAC-PapersOnLine 48-3 (2015) 294–299 Available online at www.sciencedirect.com 2405-8963 © 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. Peer review under responsibility of International Federation of Automatic Control. 10.1016/j.ifacol.2015.06.097 © 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. Keywords: Adaptive systems, Sensor systems, Multilevel systems, Process models, Business process engineering. 1. INTRODUCTION Humans are well recognized as an important element for increasing flexibility, agility, and competitiveness in future factories (cf. EFFRA, 2013) e.g., high variability or cus- tomization pressure create a demand for human-centered automation solutions. The achievement of manufacturing objectives relies on the interaction among humans and machines. For such systems, adaptation and dynamic sys- tem changes are considered important for optimal task allocation between humans and machines. Adaptation and dynamic changes can relate to physical, sensorial, and cognitive capabilities of workers. Advances in the field of human-sensing enable to observe human properties and physical traits in manufacturing situations (Teixeira et al., 2010; Patel et al., 2012) in order to align the interactions among humans and machines. In principle, various contextual factors like work task, availability of tools, physical condition of worker or tem- perature at the shop floor may influence manufacturing situations. Literature in the field of occupational psychol- ogy reveals that “the number of employees experiencing psychological problems related to occupational stress has increased rapidly in Western countries” (Van der Klink et al., 2001, p. 270). Occupational stress may result in considerable costs regarding absenteeism, loss of produc- tivity, and health care consumption. Beyond that, stress may cause high rates of tension, anger, anxiety, depressed mood, mental fatigue, and sleep disturbances at an indi- vidual level. Physical diseases or burnout may be the result in the worst case (Van der Klink et al., 2001). Conti et al. (2006) investigated influencing factors on worker stress in lean production implementations. The results indicate that job demands may cause stress for workers. For example, increasing the pace and intensity of work above the normal magnitude leads to an increased stress level. Other examples that may influence the stress level of workers are working longer than desired hours, cycle times, doing work of absent workers, feeling of blame for defect or ergonomic difficulties. Furthermore, team working, adequate task and tool support, and worker participation in process improvements have been identified to be beneficial for reducing stress. So how could we measure stress of humans and adapt workplaces accordingly? Basically, stress can be measured by applying (i) psychological questionnaires (ex post to stressful situations) or (ii) by measuring physiological * Department of Business Information Systems - Communications Engineering, Science Park 3, Altenberger Straße 69, 4040 Linz, Austria (e-mail: matthias.neubauer@jku.at) ** Department of Business Information Systems - Communications Engineering, Science Park 3, Altenberger Straße 69, 4040 Linz, Austria (e-mail: florian.krenn@jku.at) *** MA Systems & Control Limited, 89 High Street, West End, Southampton, SO30 3DS, United Kingdom (e-mail: dennis.majoe@inf.ethz.ch) Abstract: Today’s production companies face a variety of challenges arising from increasingly complex and dynamic environments. The advent of pervasive computing has enabled companies to continuously adapt to these changing manufacturing situations. In factories of the future the worker and his or her well-being is seen as a crucial part of manufacturing situations. These human factors have to be considered in order to achieve sustainable organizational success. Due to advances in the area of wearable sensors, sensing human properties within a manufacturing setting is technically feasible. Sensing human properties, such as the level of comfort or stress, provides additional information. This allows for continuous adaptation of the manufacturing system behavior based on human needs. In this paper, human-aware modeling and the execution of production processes incorporating human properties are illustrated. This is done by applying the Subject-Oriented Process Management approach in an application scenario. Furthermore, the architecture and the application of the designed system are described with respect to human stress level and dynamic system adaptation. Towards an Architecture for Human-aware Modeling and Execution of Production Processes Matthias Neubauer * Florian Krenn ** Dennis Majoe ***