Agent-Based Modeling of Mass-Collaborative Product Development Processes

摘要

Mass-collaborative product development refers to a paradigm where large groups of people compete and collaborate globally to develop new products and services. In contrast to the traditional top-down decomposition-based design processes, the primary mechanism in mass-collaborative product development is bottom-up evolution. Hence, the issues underlying mass-collaborative processes are fundamentally different from those in traditional design processes. For example, instead of determining the best sequence in which activities should be carried out, the emphasis is on developing the right conditions under which product evolution can be fostered. Existing research on product development is primarily focused on top-down design processes. The evolutionary nature of mass-collaborative product development has received very little attention. Specifically, computational models for these processes have not been developed. In this paper, a step toward understanding the fundamental processes underlying mass-collaborative product development using a computational model is presented. The model presented in this paper is based on an agent-based modeling approach, which allows the modeling of the behavior of different entities within a product development scenario and the study of the effect of their interactions. The model captures the information about (i) products as modules and their interdependencies, and (ii) the participants involved and their strategies. The benefits of the agent-based model in understanding mass-collaborative product development are shown using a simple product model. The following aspects of the product development processes are studied: (a) the rate of evolution of the individual modules and the entire product, (b) product evolution patterns and the effect of the number of participants, (c) the effect of prior work on product evolution, (d) the evolution and distribution of participants, and (e) the effect of participant incentives. The agent-based modeling approach is shown as a promising approach for understanding the evolutionary nature of mass-collaborative product development processes.