Towards Software Evolution for Embedded Systems Based on MAPE Loop Encapsulation

Abstract

Software evolution is an essential activity that adapts existing software to changes in requirements. Because of recent rapid requirements changes, systems are strongly required to evolve even if the target systems are embedded systems, whose implementation code is generally hard to be changed. This paper discusses the feasibility of applying self-adaptation mechanism for software evolution. We use the MAPE loop mechanism to evolve embedded systems without changing code inside the existing systems. This paper also reports preliminary results that we experimentally evolved a cleaning robot by following our approach. Our demonstrations show a part of additional behaviors as the results of software evolution that makes the cleaning robot possible to move obstacles. We also discuss the future directions of software evolution for embedded systems with the self-adaptive mechanism.