Focus on…Marielle Malfante (GIPSA-Lab) and the Vosica project

on the August 28, 2018

Marielle Malfante is a PhD student at the GIPSA-Lab laboratory (UMR of Grenoble Alpes University and CNRS). The team she is involved in obtained funding from the Grenoble Alpes Data Institute for the VOSICA project linked to her thesis.
The VOSICA project has been developed in the framework of Marielle Malfante’s thesis supervised by Mauro Dalla Mura and Jérôme Mars from the GIPSA-Lab laboratory Sigmaphy team and supported by LABEX OSUG2020 and DGA/ MRIS. The aim of Marielle Malfante’s thesis is to develop automatic methods of classification in natural environments. She is working on underwater acoustics and monitoring of volcanoes through the automatic classification of signals.

Monitoring volcanoes is essential for the security. Sensors are installed on the surface of volcanoes to monitor vibrations. Few experts are able to associate these vibrations to natural phenomena and predict volcanic eruptions. When a volcano is continually monitored, a huge amount of data is produced. The analysis of the acquired signals is no longer possible to be done manually.
The goal is now to detect and classify automatically and in a real time the different types of recorded signals. It will be a tool for decision support for the security recommendations made to local authorities.

Marielle Malfante uses machine learning to create an algorithm that will precisely describe signals and classify them: thanks to previous data, it will analyze and recognize future signals. She had to deal with three difficulties. Firstly, the data used to build an analysis model for a volcano are recorded on shorter time periods compare to the volcano life time. Those data are therefore hardly representative of the range of behavior displayed by the volcano. Secondly, if a new phenomenon appear, it won’t be recognize by the automatic analysis since the analysis model is created from previously recorded data. Thirdly, each volcano is characterized by a peculiar set seismic activity and so the tool for the automatic analysis need to be adapted.

VOSICA project with the Grenoble Alpes Data Institute funded a mission for M. Malfante and R. Al Warda (Indonesian student from Gadjah Mada University enrolled in MEES Master @UGA) to develop and deploy the tools for the automatic analysis of seismic signals of Merapi, the most dangerous volcano at Java Island, on the servers of the Indonesian BPPTKG observatory. The PhD student was pleased to go there to realize the application part of her thesis.


All photos: © GIPSA-Lab / Marielle Malfante



Published on August 30, 2018