PeNeTo; A Petri Net Simulator for Fast Safety and Quality Analysis and Cost - Prediction

PeNeTo is Petri-net tool for modeling, visualizing and analysing extended generalized stochastic Petri nets, which is being developed at DaimlerChrysler Research. It contains powerful modeling and parameterization functionalities, analysis by both discrete-event simulation and numerical techniques, and a graphical visualization of the token game. It is used with discrete-event simulation for various system studies at DaimlerChrysler; numerical analysis via the method of supplementary variables and phasetype distributions is currently under development. The goal of this paper is to present PeNeTo and some of the areas in which it is used. We first describe the environment in which PeNeTo is being developed and applied. From this we derive requirements for the tool's functionality, which are then described. Part of this functionality is demonstrated using an example application. In the final section, the current state of PeNeTo is described, together with plans for its further development. GOALS AND APPLICATIONS OF PENETO Modern road vehicles contain increasingly complex and interconnected systems, with everincreasing levels of connectivity and interaction. This immense growth in complexity makes it more and more difficult to make accurate predictions about such systems with a reasonable amount of effort and time. For this reason, research and development in the automotive industry needs easy-to-use and powerful analysis tools. One example of this is to build a formal model of a system which can be easily analyzed by computer; PeNeTo is one such tool for the modeling and analysis of stochastic Petri nets. PeNeTo is designed for a broad range of applications, including safety, quality and cost analysis. Safety analysis is concerned with the design of safe automotive systems, including their interactions with humans. Safety analysis makes use of reliability, availability and safety information such as load and throughput of system components and probabilities for critical states. Quality analysis is concerned with building models at the component, system and vehicle levels for obtaining information on product quality. One example of this is comparing the behavior of systems built using different components. Cost analysis allows, for example, the calculation of price models for maintenance and service packages which depend on the level of service offered, as well as the age and mileage of the vehicle. In addition, predictions can be made for costs incurred to the manufacturer during the vehicle's warranty period which are caused by quality problems. In all cases, the goal is that these analyses can be performed as fast and as comfortably as possible, while covering a large number of variants, in order to study and compare different real-life scenarios. PeNeTo is designed with a broad functionality; it contains a large number of the various extensions to stochastic Petri nets that have been proposed, including immediate transitions, guard functions, user-defined rewards, place capacities, arc multiplicities, marking-dependent firing rates, single-server and infinite-server transitions, and age and enabling memory policies.