Design Methodology of Microservices to Support Predictive Analytics for IoT Applications.

Area of Science:

• Computer Science
• Data Science
• Software Engineering

Background:

• The Internet of Things (IoT) is rapidly evolving with advanced data handling capabilities, but faces challenges in integrating heterogeneous data and providing adaptable services.
• Existing IoT analytic systems struggle with efficiency, maintainability, and scalability due to monolithic architectures and diverse data sources.

Purpose of the Study:

• To propose a design methodology for an efficient and scalable IoT analytic system.
• To embed analytic capabilities within modular microservices to support adaptive IoT applications.
• To address the challenges of heterogeneous data and evolving requirements in IoT environments.

Main Methods:

• Developed a design methodology integrating analytic capabilities into modular microservices.
• Created algorithms for analytic procedures to support the proposed model.
• Utilized Web Objects for IoT resource virtualization and semantic data modeling for interoperability.

Main Results:

• The proposed design facilitates efficient and scalable services for adaptive IoT applications.
• Modular microservices enhance the maintainability and adaptability of IoT analytic systems.
• Semantic data modeling improved interoperability across heterogeneous IoT systems.

Conclusions:

• The microservices-based approach offers an efficient and scalable solution for IoT analytics.
• The design methodology effectively addresses the challenges of heterogeneous data and dynamic requirements.
• The prototype implementation validates the proposed design for adaptive IoT applications.