MDD4CPD: Model Driven Development Approach Proposal for Cyber-Physical Devices

Abstract

Context: the Internet of Things (IoT) is a paradigm that provides an ecosystem for a fast-growing quantity of connected devices, also defined as cyber-physical devices. Problem: the creation of Internet of Things solutions is fairly complex, having to integrate and communicate between sensors, devices, and larger systems, presenting many technical challenges not present in the same magnitude as other paradigms. One of the most affected segments is the development of cyber-physical devices. Much of its development energy is spent on the connecting and efficacy of these devices, often overlooking the future impacts of the proposed solution, caused by a lack of software quality. SI Theory: this work was designed under the Design Theory, currently found in the Justify/Evaluate stage’s refinement process. Methodology: model-driven development (MDD), a software development methodology that allows the generation of software solutions through abstract models, which aims to facilitate development, bringing the abstract solution closer to the problem’s implementation. In this work, a custom MDD approach for cyber-physical devices is presented. Results: the proposed approach and tool were able to generate C++ source code for Arduino devices, the generated source code performed better than control when compared using OOP specific software metrics. Contributions: this study presents an MDD approach for cyber-physical devices, where the innovative models are based and coupled to the hardware specifications, being extensible and flexible.