Atsafack Lab

Atsafack Lab

AI-driven insights for smarter hydropower maintenance

See More
profile

Hydropower AI

A large dam structure spans across a body of calm water, with the dam itself featuring industrial elements like metal railings and a control tower. The surrounding area consists of rugged terrain with dense greenery and pine trees covering the hills in the background. Overhead, the sky is clear and blue, with visible electrical pylons atop the hills.

LSTM Models

Developed deep learning LSTM models to predict equipment failures before they happen, reducing downtime.
A grey processor chip with the letters 'AI' prominently displayed in blue, set against a subtle background with faint outlines of a map.

Smart Sensors

Implemented IoT smart sensors for continuous monitoring of hydropower plant components, enabling proactive maintenance.
A vintage typewriter is displaying a sheet of paper with the words 'EDGE COMPUTING' typed in bold letters. The typewriter features a dark green body and metal components, set against a soft white background.

Anomaly Detection

Created algorithms to detect unusual patterns in operational data, alerting operators to potential issues early.
A large rusted industrial machine with intricate gears and mechanisms is set against a clear blue sky. The metal parts are weathered, showing signs of aging and use, with a mix of brown and red oxidation visible. The machine's design suggests it is used for heavy-duty mechanical work or construction.

Predictive Maintenance

Analyzed energy output trends to optimize performance and support sustainable hydropower management.
young-beautiful-african-woman-student-headphones-sitting-cafe-reading-book-education-learning
Publications

About Me

Ms. Blondelle Melina Atsafack is a researcher and data analyst specializing in Internet of Things (IoT) and Machine Learning applications. Her work focuses on developing intelligent systems for predictive maintenance and real-time monitoring, with particular emphasis on hydropower plant facilities. She is passionate about leveraging data-driven approaches to improve the reliability, efficiency, and sustainability of industrial operations, while advancing innovative solutions for energy and smart infrastructure in Sub-Saharan Africa and beyond.

My Work

My work focuses on the development of IoT-enabled predictive maintenance systems for hydropower plants. I design intelligent architectures that integrate sensor data acquisition, real-time monitoring, and deep learning-based anomaly detection models, particularly LSTM Autoencoders and domain-adaptive techniques. Through multivariate time-series analysis, I aim to improve fault detection accuracy, reduce unplanned downtime, and support data-driven maintenance decision-making in hydraulic turbine systems.