Create an improved ADAS based on the prediction of traffic participants behaviors in order to warn a driver early enough to fully avoid a collision.- Problem: Too many people died in car accidents. Most accidents happen due to a late reaction from the driver.- Existing solutions: ADAS aim at arlerting a driver shortly before a collision. In most cases, the driver has time to reduce its velocity but the impact is not fully avoided. Drivers are warned too late.- New solution: Drivers should be warned earlier to fully avoid a collision. A solution is to use the prediction of traffic participants.- Goal: Develop a generic and adaptable framework based on the combination of multiple context-based models in order to predict future surrounding traffic participants behaviors and alert a driver earlier about a risk of collision.Develop knowledge in classification methods, machine learning algorithm, scene modelling, behaviour prediction and project management.Publications:- S. Bonnin, F. Kummert, and J. Schmuedderich, A Generic Concept of a System for Predicting Driving Behaviors, in Proc. IEEE Intelligent Transportation Systems, 2012, pp. 1803-1808.- S. Bonnin, T. H. Weisswange, F. Kummert, and J. Schmuedderich, “Accurate behavior prediction on highways based on a systematic combination of classifiers,” in Proc. IEEE Intelligent Vehicles Symposium, 2013, pp. 242–249.- S. Bonnin, T. H. Weisswange, F. Kummert, and J. Schmuedderich, “Pedestrian Crossing Prediction using Multiple Context-based Models,” in Proc. IEEE Intelligent Transportation Systems, 2014, has received the Best Student Paper Award.- S. Bonnin, T. H. Weisswange, F. Kummert, and J. Schmuedderich, General behavior prediction by a combination of scenario-specific models,” IEEE Transactions on Intelligent Transportation Systems, vol. in press, pp. 1–11, 2014.

Final year placement at Honda Research Institute in Germany. Work on a Vision system for car domain to detect the motion of the different objects of the environment.Estimation of Motion and Odometry for Automotive and Robotic Scenarios. Compensate the EgoMotion to compute the Optical Flow and estimate the Proper Object Motion in the scene. How to optimize the quality of the EgoMotion compensation to compensate the car motion due to the road.

Internship in England in Centre for Robotics & Intelligent Systems,SOCCE at Plymouth University. I worked on a autonomous mobile robot designed to serve drinks at parties. I worked on the Vision system based on stereovision, the motion detection to locate the people, human tracking in the room and bring them the drinks after estimating the distance to follow.

Integration of humidity and temperature sensors, involving converting analogue information into a digital output to be displayed on an LCD. Responsible for the design, sourcing of parts and construction of the necessary hardware and microcontroller software. Proteus was used to develop the electronic circuit. The microcontroller was a PIC16F877 by Microchip programmed in PIC.