Proceedings of TMCE 2012, May 7–11, 2012, Karlsruhe, Germany, Edited by I. Horváth, A. Albers, M. Behrendt and Z. Rusák  Organizing Committee of TMCE 2012, ISBN ---- 1 BULK SIMULATION OF USING INFORMATION-INTENSIVE PRODUCTS AND PRODUCT-SERVICE SYSTEMS: FORMAL UNDERPINNINGS Wilhelm Frederik van der Vegte Faculty of Industrial Design Engineering Delft University of Technology w.f.vandervegte@tudelft.nl Imre Horváth Faculty of Industrial Design Engineering Delft University of Technology i.horvath@tudelft.nl ABSTRACT In virtual usability testing, simulating the workings of products is typically a computation-intensive pro- cess. For interactive setups, researchers have put much effort in realizing real-time physics simulations that can keep up with human subjects. But what if faster-than-real-time simulations would be possible? Human subjects cannot synchronize with such simu- lations, but in non-interactive setups, where algo- rithms approximating human (re)actions are used as surrogate subjects, new opportunities would emerge. Fast simulations allow investigating long-term use processes (weeks, perhaps even years of use), or per- forming massive batch testing with different design parameters or different user characteristics, and thus designers can potentially identify a larger range of usability problems. We hypothesized that infor- mation-intensive (or ‘smart’) products that not so much depend on complex physics phenomena offer an excellent opportunity to realize faster-than-real- time simulations. Building on previous work on non- interactive simulation, this paper gives the concepts and the formal theory of a new approach based on a further simplified reasoning model of human-artifact interaction, and exploiting opportunities to use sim- plified physics models. We also present a first demonstrative example including a provisional vali- dation of the claim that a performance much faster than real-time can be achieved. In addition to ena- bling fast large-scale or ‘bulk’ simulations, our ap- proach allows extension towards product-service systems, which are considered relevant in the context of smart products. Future work includes adopting cognitive simulation abilities, automated scenario generation and scheduling batch simulations. KEYWORDS Smart/information-intensive products, use process simulation, bulk simulation, timed hybrid automata, product-service systems. 1. INTRODUCTION With the rise of smart systems and ubiquitous com- puting, information-intensiveness of products in- creases and users are challenged – possibly even overloaded – by the expanding quantity of options and possible interactions. The use process may be- come even more complicated if the product is part of a product-service system with multiple networked products and multiple participants, which is a typical setting where smart products are used [1]. These drivers of complexity can rapidly escalate the num- ber of possible variations of user-operation sequenc- es, and for designers it becomes increasingly difficult to foresee all possible outcomes, which might include unacceptable performance, failure, and even fatali- ties. Our novel method to predict possible outcomes of use during design, and thus allow anticipation be- fore the product/system is realized, aims to help solv- ing this problem. In previous work [2,3] we showed how complex in- teraction processes can be virtually tested by control- ling conventional engineering simulations with use scenarios. Virtual testing can be done completely on a computer without deploying human subjects or physical prototypes. This allows early risk identifica- tion during design. A scenario describes ‘a way to use a product’. By unifying multiple ways of use, scenario bundles address the fact that most products are used in multiple different ways, and also multiple times.