TL;DR
This paper tackles the challenge of enabling domain experts to easily develop applications for the Internet of Things (IoT). They propose a domain model that captures the essential concepts and relationships within the IoT, aiming to provide a common understanding and a foundation for future development frameworks.
What Problem Does This Paper Address?
The Internet of Things (IoT) is rapidly expanding, connecting physical objects to the internet and creating new possibilities. However, a significant hurdle is making it easy for domain experts (like healthcare professionals or city planners) to create IoT applications without needing deep computer science expertise. Current approaches often require specialized knowledge, hindering broader adoption. This paper aims to bridge this gap by defining a clear conceptual model for IoT applications.
Methods
The authors developed their domain model through a two-step process:
- Survey of IoT Applications: They conducted an extensive survey of existing and proposed IoT applications from research and industry. This helped them identify common behaviors and patterns.
- Domain Modeling: Based on the survey, they extracted key concepts and their relationships. They used the CRC (Classes, Responsibility, Collaboration) technique and a graphical representation (UML-like) to define the model. The model categorizes concepts into "Traditional Internet Concepts" (like software components, computational services, storage services, and end-user applications) and "Thing-Oriented Concepts" (like entities of interest, phenomena, resources, raw data, actions, devices, drivers, and commands).
Datasets
Not discussed in the paper. The paper focuses on a conceptual model derived from existing literature and industry trends rather than empirical data from specific datasets.
Key Insights and Ideas
- The Need for a Common Language: The paper emphasizes that a shared vocabulary and conceptual model are crucial for effective communication and development among domain experts, researchers, and system programmers in the IoT space.
- Invariant Properties are Key: The domain model captures the fundamental, unchanging aspects of IoT concepts and their relationships. This invariance makes applications more resilient to technological shifts.
- Separation of Concerns: The proposed model implicitly suggests a layered approach (resource, service, end-user) that helps in organizing and understanding the different functional aspects of an IoT application.
- Bridging the Physical and Digital: The "Thing-Oriented Concepts" are vital for representing the physical world's entities and their observable properties (phenomena) and how they interact with the digital realm through sensors and actuators.
Key Results
The primary outcome of this work is the proposed domain model itself, which is presented graphically in Figure 1. This model aims to:
- Provide a common understanding of IoT concepts.
- Model invariant properties of IoT domain concepts.
- Enable modular design by clearly identifying entity capabilities.
- Facilitate the integration of real-world things, software, and network models.
The authors mention that they have used this model to successfully represent three types of IoT applications: shared book reviews, smart plants, and room HVAC maintenance.
Limitations and Future Work
The authors acknowledge that their work is an initial step and that more research is needed. Their future work includes:
- Expanding the survey of IoT applications to further refine the domain model and capture a more exhaustive set of behaviors.
- Elaborating on interaction paradigms within IoT applications, specifically detailing the "communicate-with" association between software components (e.g., message passing, publish-subscribe).
- Developing a high-level programming model, likely a Domain Specific Language (DSL), to provide domain experts with a user-friendly framework for IoT application development.
Summary updated