This essay contributes to solving two issues: (1) clearly defining Non-Functional Requirements (NFRs) and Functional Requirements (FRs), and (2) simulating on a practical IoT platform. Related to the first problem, we always feel annoyed with many irregular conditions and non-requirements for programming physical systems. Particularly, they sometimes cause cross-cutting concerns problems. Thus, we cannot concentrate on mainstream behavior. Regarding the second problem, the platform must deal with IoT problems. Modern physical systems are integrated with the Internet of Things (IoT), which connects multiple devices, sensors, and cloud services. As a result, these systems may face problems such as latency, race conditions, deadlocks, and more.

To solve these problems, we propose a robot software development environment called CPy4NFR. For the first problem, we draw NFRs in feature models, and CPy4NFR generates layers of Context-Oriented Programming (COP). The programming language is called CPy, which is an extension of Python. Through this process, the relation between non-functional requirements and COP is clarified, and the cross-cutting concern problems are solved. Regarding the second problem, CPy programs with NFRs are executed on Hakoniwa. Hakoniwa deals with IoT problems, provides APIs for simulator environments such as Unity and Unreal Engine, and supports APIs for physical robot systems. In this essay, we apply CPy4NFR to develop a drone system with changing behavior at runtime. Finally, we discuss two problems and the proposed development environment.