In 2024, Hornos and Quinde conducted a very thorough methodological review that is to the point of stating the lack of a common development method for the Internet of Things (IoT)-based systems, which is the ongoing problem in as the situation is far from settled. The situation is complicated by the fact that IoT applications are gaining popularity very rapidly in different important areas like healthcare, smart cities, transportation, and industrial automation. At the same time, the integration of diverse hardware, software, networking, and intelligent decision-making systems has become more difficult. The text argues for considering the software engineering discourse by questioning whether the existing development methodologies—mostly carried over from the traditional information systems—are suitable for the complex and changing character of IoT environments.
The authors of the article see IoT platforms as highly complex social-technical systems that not only pack a wide range of devices but also control their communication amongst each other and the entire system at a very high speed, thanks to the computing infrastructures in the cloud, fog, and edge. They argue that the integration of hardware configuration, interoperability, security, and lifecycle management still poses a challenge to the use of traditional software development approaches in IoT systems.
Moreover, the document points out that IoT development is faced with heterogeneity and interoperability, non-functional requirements management, reliability, and dependability, user-centric design, security and privacy, and AI integration constraints as the main obstacles. Based on these challenges, the authors put forward a set of research directions, demanding integrative, flexible, and standards-aligned methodologies that unite model-driven engineering, agile principles, AI support, and formal verification methods.
The article is logically structured, starting with the conceptual background and then proceeding to the methodological analysis, identification of challenges, and suggesting future directions for research. The authors’ argument that there is no existing methodology that completely covers the entire IoT lifecycle is very well supported by the evidence provided through comparison.
However, the critique at times comes close to being a generalisation of the methodology. Some of the particular methodologies are assessed largely by their concepts, without recognising the domain-specific accomplishments that could be the case when the methods are applied in a narrow context. Therefore, the argument might miss the point about the suitability of some existing methodologies in the specified context.
Hornos, M. J., & Quinde, M. (2024). Development methodologies for IoT-based systems: Challenges and research directions. Journal of Reliable Intelligent Environments, 10, 215–244. https://doi.org/10.1007/s40860-024-00229-9