In sub-Saharan Africa, efforts to ensure food security and sustainably transform agriculture often face fragile transport systems. Deteriorated rural roads and a lack of accident prevention measures jeopardize the transport of agricultural products, particularly for heavy vehicles. Accidents, particularly truck rollovers, lead to significant post-harvest losses, increased logistics costs, and economic insecurity for producers. This paper presents the objectives of an ongoing doctoral research project aimed at developing a systematic tool for detecting the risk of heavy vehicle rollovers on the roads of developing countries, with a focus on Benin. This system, based on an interdisciplinary approach combining automotive engineering, data analysis, and artificial intelligence, will identify critical risk areas, predict dangerous scenarios, and alert drivers or fleet managers in real time. The innovation lies in the integration of road data, vehicle dynamics, and machine learning models to anticipate accidents. In this context, modeling heavy goods vehicle dynamics is the first step in developing the tool. The goal is to contribute to securing transport corridors, reducing economic losses, and promoting smooth and reliable access to markets essential conditions for sustainable economic growth and improved food security. Keywords: Road safety, Agricultural transport, Artificial intelligence
AMOUSSA YOUSR, SEMASSOU G. CLARENCE, AMADJI T. ARMEL