High-level open evolvable systems design by process-oriented modeling

Abstract

Open Evolvable Systems' design requires a methodological [1] and conceptual paradigm different from the conventional software design. Evolvable Systems' research [2, 6, 16, and 17] has established itself as a new research field, but the content is more domain-oriented than universal. Consequently, major contributions toward substantiation of that universal methodological and conceptual paradigm are yet to come. In this paper we present a new perspective and method for the general-purpose design of Evolvable Systems. The paper presents the attributes of the Evolvable Systems and discusses the distinction between Evolvable Systems' and conventional software design as well as the methodological ramifications. We pose and address the question of what is an efficient methodology for designing a system for which we do not know the boundaries? We present our version of Process-oriented Modeling as the key method in the high-level design of Evolvable Systems and show its utilization in implementation of one modeling case of a complex Evolvable System, the DNA replication process. We also present the dynamic aspects of the design process management and pre-code verifications in the framework of Quantified Controls and Simulations.