Back to the listing
Mohamed B.

Embedded Software Developer

345 dollar
Freelancer
3 years
Paris, FRANCE
Analysis C# C/C++ Deep learning Linux Machine Learning Matlab Python

My experience

More

AltranFebruary 2020 - Present

Description : Trailer Assist
Trailer Assist consists in developing a method of measuring the angle between the trailer and the vehicle from the rear view camera (2D image). This measurement system is mainly used in the parking assistance system of a vehicle with a trailer, it must therefore provide an estimate of the angle value in real time.

OBJECTIVES:
• State of the art study: trailer assist technology offered by car manufacturers such as VOLKSWAGEN, Ford, Audi and Land Rover,
• Construction of a database (Images, Videos),
• Data preprocessing (Images, Videos),
• Development of an algorithm based on artificial intelligence to measure the angle between the trailer and the vehicle in real time from 2D images generated by the rear camera of the vehicle,
• Embedding the optimal algorithm in an embedded target.

REALIZATIONS:
• Acquire and annotate the database for training and validation,
• Code a basic CNN model in Python,
• Code and test the end to end solution with EfficientNet capable of estimating with an accuracy of 1 degree the value of the angle between the trailer and the vehicle from a 2D image,
• optimize the code for real time in python (

More

AltranSeptember 2019 - February 2020

Description : Trajectory-based ego-positioning
The project is part of: Enhanced perception for the realization and within the scope of Work Package "E-Horizon & Localization", of two activities called "Trajectory-based ego-positioning" and "V2X for localization.

OBJECTIVES:
• Requirements analysis,
• Algorithmic exploration,
• Development, integration and validation of application software.

REALIZATIONS:
• Analysis of the requirements of the "trajectory-based ego-positioning" functionality inside the precise location component,
• Algorithmic exploration of the functionality of "trajectory based ego-positioning" for the centimeter localization of the ego including also a study of the state of the art,
• Comparative analysis of algorithmic performance based on simulation,
• Matlab development of the chosen algorithm.

TECHNICAL ENVIRONMENT:
• Matlab script, gitlabee, JIRA.




More

AltranFebruary 2019 - July 2019

Description : Multi-sensor data fusion and synchronization for an autonomous vehicle

OBJECTIVES:
• Multi sensor data fusion for autonomous vehicle,
• Responsible for data synchronization,
• State of the art of existing methods and choice of methods used,
• Programming algorithms in C ++ under ROS,
• Testing and optimization of algorithms.

REALIZATIONS:
• Use of a centralized fusion architecture (preprocessing, data synchronization, spatial alignment, time alignment and data association),
• Camera data preprocessing using CNN Deep Learning architecture,
• Leddar data preprocessing using mathematical algorithms in C ++,
• Time synchronization between the clock of each sensor and the clock of the PC (the host),
• Spatial alignment,
• Time alignment of all data,
• Documentation of C ++ codes using Doxygen.

TECHNICAL ENVIRONMENT:
• ROS, C++, Python, Linux, Doxygen,






More

Coficab Maroc S.A.February 2018 - June 2018

Description : Study, Automation and Supervision of a material suction system in the extrusion lines

OBJECTIVES :
• Resizing of the electrical equipment,
• Automation of the system with the SIEMENS SIMATIC S7-1200 station,
• Design of a command-control HMI.

REALIZATIONS:
• Realization of the different electrical diagrams of the extrusion area,
• Sizing and choice of different electrical equipment,
• System automation using LADDER language,
• Design of an HMI to view the status of each sensor and actuator as well as the weekly consumption,
• Evolution of the Running Time from 87.79% up to 90% for the extrusion zone.

TECHNICAL ENVIRONMENT:
• TiaPortal, X-RELAIS






More

Lear CorporationJuly 2017 - September 2017

Description : The design of a technical solution to automate the diagnosis of an Autotester of automotive ECUs

REALIZATIONS :
• Manual test of one of the components of the "AJB" Automation Junction Box, using the black box method,
• Proposal of a diagram of an electronic card to automate the AJB test,
• Realization of a Prototype card,
• Test and Validation of the electronic card.

TECHNICAL ENVIRONMENT:
• PSIM, ISIS




My stack

Machine Learning

Deep learning

Others

Electrical engineering, Artificial Intelligence, Analysis

Frameworks

OpenCV

Analysis methods and tools

Requirements Analysis, JIRA

IT Infrastructure

Linux

Technologies

Machine Learning, NumPy

Software testing

Manual testing

Languages

Python, C++, C/C++, Matlab, C#

My education and trainings

Certifications - Udemy

Speak in public - OpenClassrooms2020

Masters Degree, Embedded Systems - National School of Engineers South Alsace ENSISA2018 - 2019

Engineer's degree - Electronics, Electrical engineering & Automatic - Faculty of Sciences and Techniques of Tangier2015 - 2018

Erasmus exchange - Electronics, Electrical engineering & Automatic - National School of Engineers South Alsace ENSISA2016 - 2017

Baccalaureat, Mathematical science - Technical High School MOULAY YOUSSEF2012 - 2013