Testing tools developer for autonomous vehicles.
The objective of this internship is to develop a c++ application that enables a
drone to learn a path so that it can reproduce it autonomously.
This application has to take into account :
- Functional objectives :
a) Connect the app to the drone
b) Define a « learning path » mode where the user will guide the drone, whilst
it records the path it is taking.
c) Maintain the possibility to command directly the drone.
2)Technical requirements :
a) Use the SDK of the drone's manufacturer.
b) Develop a mobile app : possibly a hybrid that works on both Ios and Android
Other functionalities can be taken into account such as :
-Recording several paths.
-The planification of such paths.
-Visualizing recorded paths on a map.
1 Overview and understanding of the ROV activity
a) Fault location
b) Cable deburial
c) Cable gripping
d) Post repair inspection and burial
e) Post lay inspection and burial
2 Meucci ROV performances
a) assessment of the operational performances compared to theoretical
performances
b) characteristic of the umbilical
c) characteristic of the crane and its performances
3 Analysis of clients requests
a) contractual technical specifications and performances in maintenance
contract ( Mecma)
b) contractual technical specifications and performances in maintenance
contract Sea Me We 5, OTE, Junior
4 control station and umbilical interface
a) Theoretical study to adapt another ROV on the same umbilical and same
crane
Designed in Java a software to facilitate risk management for the company.
(The software enables the company to develop a risk matrix and estimate the
gravity of scenarios that could affect the company)
In a team of 4, I played an active role in the project from the planning, to the
conception and implementation of the solution.
We were in direct contact with the client to improve the software, discuss
technical issues and provide technical support.
Developed in javascript a prototype of 3d Augmented Reality software that
runs on a web browser.
I worked in a team of 4 students.
Augmented reality can usually be divided into 4 parts:
-The image recognition
-The calculation of the homography matrix
-the tracking of corners
-the 3d rendering
We implemented the last 3 parts.
We used the Jsfeat library to obtain the homography matrix that gives the
position of the camera regarding the position of the object, as well as to track
the image.
The 3d rendering was implemented with three.js. We created a virtual camera
thanks to the homography matrix and displayed a cube on the webcam image.
Prepared parcels in a team of 4 and reorganized a part of the warehouse in
order to facilitate the search of products.
Translated files from English to French, and worked in open-space in a
multicultural environment.