Prototyping a demo Web Application for ML Models, Technilog

About Technilog

Technilog is a service company specializing in the development of software solutions for the integration of connected industrial objects or IIOT (“Industrial Internet of Things”). Technilog offers two flagship software solutions: Dev I/O, a remote management and supervision software that unifies and processes data from connected equipment, and Web I/O, a platform for the supervision and control of connected objects across various platforms.

Source: Technilog

Context

As part of the evolution of IoT solutions offered by Technilog, a new version of their flagship product, Web I/O, has been developed. This version integrates advanced features, including artificial intelligence (AI) tools. To illustrate these tools, a web application was created to demonstrate the use cases and capabilities of this new version.

Achievements

An initial version of this demonstration site was developed by the previous intern. In this version, data qualification functionalities and variable anomaly detection were implemented. Additionally, the necessary infrastructure for web hosting of the application was set up. My role was to add new features and participate in the design meetings for the web pages:

• Anomaly Detection: Based on machine learning from historical data, this feature determines if a value behaves unusually.
• Forecast Simulation: Similar to anomaly detection, this feature uses machine learning from historical data. From this learning and previously obtained values, a model can be determined to anticipate future values.
• Impact Study: Allows evaluating the impact that different variables have on each other, as well as the relationships between these variables, providing an in-depth understanding of the mutual influences of these components.
• Quality of Life Improvements: Addition of authentication features and model updates.
• Deepening Knowledge: In-depth study of various artificial intelligence models.