During the Crunch Industry Camp, a 48-hour non-stop engineering event that brings together volunteer students from the UTBM and companies like Orange to collaborate on real-world problems, I had the opportunity to contribute to an important project proposed by Orange. The event aims to provide students with an opportunity to work on real-world engineering projects and develop critical skills such as problem-solving, teamwork, and communication.

The problem of transportation pollution, particularly from trucks, is a pressing concern that requires attention. One key step towards improving the efficiency of goods transportation by trucks is to accurately estimate their circulation. To this end, Orange has put forward a proposal for developing a publicly accessible database of truck circulation on French roads, utilizing the government's highway cameras. We were in charge of developping an autonomus system to feed that database.

Solution

The videos were captured at regular intervals, typically every 30 seconds for 15 minutes (this is a limit imposed by the government), and provided us with critical data that we needed to analyze to extract relevant information about truck circulation. To analyze the videos quickly and accurately, we used computer vision techniques and AI algorithms, specifically OpenCV. We developed an automated system that could extract relevant data, such as the number of trucks passing through, the time of day (by one-hour intervals), and the direction of travel, from the videos captured by the highway cameras. To ensure compliance with GDPR guidelines, we deleted the videos as soon as they were analyzed, retaining only anonymous figures. This process helped to protect the privacy of individuals captured in the videos while providing us with the data we needed to estimate truck circulation accurately. It should be noted that the data can only be estimated due to the shortness of the videos and the time interval between them, and a mean should be established from the different videos of each hour to ensure privacy.

Our team worked tirelessly to develop an automated system that could quickly analyze the videos and extract relevant data. We tested and refined the system until it was accurate and efficient enough to be used on a large scale. As part of the project, we also developed a user-friendly graphical interface that allowed stakeholders to interact with the data easily. This interface made it easy for anyone to access and analyze the information, allowing stakeholders to make informed decisions about transport and logistics.

Furthermore, as we had to convince a jury that our project was better than the others, we received coaching from several professionals on how to present our project, emphasizing oral and visual communication to impress the jury, visitors, and opponents. These skills are essential to my future as an engineer, as I will be required to impress clients or investors with products or projects.

Conclusion