Teaching & Popularization

Gabriel Kasmi



Teaching

Interns Supervision

February-August 2025: End of study internship of Sarah Mehiyddine

In the context of the French and European energy transition, this internship aimed to analyze the determining factors for the adoption of residential photovoltaic systems, particularly in a regulatory context evolving towards self-consumption. The main objective was to understand how economic incentives, notably feed-in tariffs, influence individual adoption decisions.

Research Article in Preparation

A scientific article is currently being written, based on the analysis of granular French data covering the period 2005-2024. This article uses detailed installation data provided by RTE and Long-term socioeconomic data from Piketty and Cagé (2023) historical database to model the profitability of PV installations. We account for the specificities of injection configurations (full vs. surplus). We propose a two-step analysis where we first assess the impact of profitability on adoption, and then focus on the determinants of profitability itself.

Our key contributions are to show that in the current context, profitability depends primarily on electricity prices rather than feed-in tariffs. This effect becomes more pronounced as the self-consumption rate increases (the threshold appears to be around 60-70% self-consumption). We also evaluate the installed potential that could be achieved with rooftop PV in self-consumption with surplus by 2035.

This work challenges the idea that recent reductions in feed-in tariffs will necessarily slow down photovoltaic deployment. On the contrary, it highlights the growing importance of self-consumption, enhanced by technological progress in storage and flexibility.

This analysis provides policymakers and network operators with key elements to:

Extensive Literature Review

A comprehensive literature review was conducted and published, examining the key factors influencing the adoption of residential rooftop photovoltaic systems. This review adopts both macroeconomic and microeconomic approaches, identifying all determinants, particularly socioeconomic and financial ones, that guide individual decisions. The literature review is available here.

Questionnaire on Individual Determinants

A questionnaire was developed to identify and analyze individual motivations behind the adoption of residential photovoltaic systems. This survey captures not only economic factors but also social and environmental factors that influence household decisions.

March-September 2022: End of study internship of Yann Trémenbert

Photovoltaic (PV) energy grows at an unprecedented pace, which makes it difficult to maintain up-to-date and accurate PV registries, which are critical for many applications such as PV power generation estimation. This lack of qualitative data is especially true in the case of rooftop PV installations. As a result, extensive efforts are put into the constitution of PV inventories. However, although valuable, these registries cannot be directly used for monitoring the deployment of PV or estimating the PV power generation, as these tasks usually require PV systems {\it characteristics}. To seamlessly extract these characteristics from the global inventories, we introduce PyPVRoof. PyPVRoof is a Python package to extract essential PV installation characteristics. These characteristics are tilt angle, azimuth, surface, localization, and installed capacity. PyPVRoof is designed to cover all use cases regarding data availability and user needs and is based on a benchmark of the best existing methods. Data for replicating our accuracy benchmarks are available on our Zenodo repository DOI , and the package code is accessible at this URL.

The main tasks of the internship included an extensive literature review of the existing works in the field of extracting characteristics of PV systems, a benchmark of these methods and an implementation of new methods (e.g. Theil-Sen regression). All methods were extensively compared both in terms of accuracy and in terms of computational cost on BDAPPV. Yann put a particular emphasis on the environmental impact of his coding practices and the methods that he implemented. The idea to discuss the environmental impact of DeepPVMapper (appendix A of the thesis manuscript) was brought by him. This internship was co-supervised with L. Dubus.

Supervision of Group Projects

Since 2022: Applied statistics projects at ENSAE

About 20h of teaching load to a group of 2nd year students of engineering degree at ENSAE Paris ("groupe de statistiques appliquées").

In 2023, global renewable energy capacity reached a record 3,870 GW, thanks to growth of 473 GW, 73% of which came from solar power. Despite this dynamism, the geographical distribution remains uneven, particularly in Africa, where the increase was only 4.6%, well short of the continent's needs. To bridge this gap, solutions such as remote sensing, already widely used in developed countries, could play a crucial role in promoting the integration of renewable energies in developing countries. The aim of this study is to propose the first application of a convolutional neural network (CNN) to the remote sensing of rooftop PV installations in Madagascar. The students will use training data from an openly available dataset to train and test their model. The students will also discover the fundamentals of machine learning and deep learning, including empirical risk minimization, model training (loss minimization, gradient descent), model architectures and implement on a GPU-based instance a deep learning model. This course gives the students a first hands-on experience with deep learning for classification and segmentation.

The project title was "Deep learning for detecting individual solar systems from aerial images". The students tasks were to carry out a short literature review, define the fundamental notions of machine learning and deep learning (empirical risk minimization, perceptron, neural network, convolutional layers, CNN) and to implement a custom-made and a ready-to-use neural network for binary classification on BDAPPV. The students leveraged PyTorch to train and evaluate the classification model. The project repository can be accessed here and the students' implementation here.

February-July 2023 : Computer science project at Mines Paris

About 20h of teaching load to a group of 1st year students of the engineering degree at Mines Paris ("projet d'informatique"). The students task was to implement a website to visualize the outputs of DeepPVMapper (in a similar fashion as this website). The main tasks were to implement an architecture with a client and a host based on Flask, to implement spatial indexing and preprocessing of the raw data to minimize the loading times and to propose some complementary statistics (e.g. the number of systems or the cumulated installed capacity per city). The project repository can be accessed here. This project was co-supervised with R. Jolivet.

Mentoring

Talks and popularization

Talks

Popularization

In Press