Posts

1. Introduction I come from a networking background. For years, my world revolved around packets, routing tables, latency, and uptime. Over time, it became clear that the boundary between IT and OT (Operational Technology) was becoming less rigid as industrial systems moved onto standard IP networks. Networks were no longer just carrying business traffic — they were increasingly responsible for transporting sensor data, telemetry, and control signals from the physical world. That realization pushed me toward Industrial IoT (IIoT) and, inevitably, embedded systems. ...

Design Science is a research methodology and problem-solving framework that focuses on the creation and evaluation of artifacts designed to solve real-world problems. Rather than studying systems passively, Design Science emphasizes active construction, experimentation, and rigorous evaluation. Its principles have traditionally been applied in information systems and software development, but they are equally relevant to modern embedded and IIoT engineering. In complex engineering domains — including embedded systems, IoT, and distributed sensor networks — the challenge is not just writing code or assembling hardware. It’s about systematically creating artifacts that solve practical problems effectively, reliably, and efficiently. ...

The scikit-learn Iris data-set consists of 3 (Setosa, Versicolour, and Virginica) species (50 samples per species, for a total of 150 samples) of the iris flower. Each sample has four measurements: sepal length, sepal width, petal length, petal width. Given these measurements a machine learning model can predict the iris specie with a high degree of accuracy. Here I demonstrate a machine learning web application using Python, Scikit-Learn machine learning library and Flask web framework. The application is then deployed on an Amazon EC2 instance. The source is on GitHub. ...