Hugo Romat

PhD student in Computer Science at INRIA

Engineer in Human System Integration


After two years of Physics and Mathematics, I entered in an Engineering School where I have studied a lot of areas of different domain (Statistic, Biology, Signal Processing, ...). I'm interested in all kinds of technological things related to human factor. The research that I lead are based on a combination of Human System Interaction and the adaptation of these systems to humans.

I also have skills in other fields like web development, computer science, mobil robotic and knowledge management as you could see in my different projects.
I like to be multidisciplinary, and to be able to work on a very large area of skill; I think this is the definition of a true Engineer.




PhD Student at INRIA / CNRS
National Research Institue of Informatic and Automatic

- Computer Science
- Interaction Human-Machine
- Big Data
- Data Visualisation
- Human Factor

2016 - 2019


Student in National School of Cognitic ENSC in Bordeaux(FRANCE)
Institute National Polytechnic (INP)

- Computer Science
- Interaction Human-Machine
- Knowledge Management
- Project Management
- Human Factor

2012 - 2015


Laval University in Quebec (CANADA)
through an exchange in the software engineering industry.
- Human/Machin Interface
- Software devlopment
- Project management
- Robotics
- Development of WebApplication
- Design and development of a platform VideoGames

2013 - 2014


Bordeaux Preparatory Cycle (CPBx)
Data Processing Option
Department of Physics and Math at the University of Bordeaux 1

2010 - 2012



Research & Development Engineer at TKM
Grenoble, FRANCE

- Data Scientist
- Big Data
- Data Visualisation
- Human Computer Interaction

NOVEMBER 2015 - MAI 2016 (7 MONTHS)

Logo UTS

Internship in UTS (University of Technology of Sydney)

- Social robotics
- Worked on the role of social cues in the concept of affordance
- Human Robot Interactions
- Human Computer Interactions
- Web Development



Internship in Diehl Aerospace
Service Cockpit Cabin System (CCS)
Frankfurt, GERMANY

- UX design
- Consulting for ergonomics interfaces
- Needs analysis, focus group and users interviews
- Prototyping of a touch interface for the FAP of the A380
- User tests on pilots and flight staff

JUNE 2014 – AUGUST 2014 (3 MONTHS)


Internship in the Chamber of Agriculture of Gironde
Service vine and wine
Bordeaux, FRANCE

- Observation/Summaries of different filtration techniques
- Proposal and development of a (Javascript / Php / Css) Web tool to optimize the preparation of wine for the bottling
- Creation of a database for assessing the performance of new filtration techniques in comparison with standard techniques

JUNE 2013 – JULY 2013 (1 MONTHS)


Volunteer for the Citadels of wine
Section Informatic

- Help to improve the actual tool which is used to read the degustation notes
- Help to use the Optic Reader and to maintain the software and hardware



Volunteer for the International Challenge of Wine

- Volunteer when grouping juries tasters
- Help juries to taste in good degustation, take care of them



Logo bestbuddies

Real Buddies (Vrais Copains)

- Sponsorship of a person with mental retardation
- Different events and meetings



Junior events manager (I2C)

- Junior events manager in the Junior-Entreprise ingénierie et Conseil en Cognitique (i2c) of the ENSC
- Organization of various partnerships with companies
- Creation of events



- Project Management
- Organisational behaviour
- Agile/Scrum

Web Programming
- Javascript/Jquery
- Angular.js/Nodes.js/d3.js
- WebGL

Software Programming (POO)
- C/C#/C++
- Python/Java
- ActionScript
- Matlab

Robotic/Signal Processing
- Filter Kayman/Bayse
- Nao/PR2/Roomba
- Kinect
- Social Robotics

Human Factor
- Usability
- User centered design
- User testing
- Modeling and prototyping

Knowledge Management
- Analysis of tasks
- Collecting Data
- Analyzing Data
- Mask Method


Robot Roomba

Application Single Page

Software Prog.

Video Game

Robot Roomba with a Kinect

My project was made with the University Laval.
My partner and I worked on navigation and localization attributes of the robot Roomba with different sensor (exteroceptive and proprioceptive).
But we used more the Kinect, which was the main sensor that we wanted to use.
To do that, we have worked on different way to interpret sensor's results with filters (Bayesian filter, Kalman filter, Particles filter …) and be able to interpret correctly the results.
At the end of our project, the robot was able to avoid obstacles with the Kinect sensor, was able to perform localization task in a simple environment and to update the map he made with what it sees.

Technology Used : Matlab/Linux


Development of an online auction and shopping with a Single Page WebApplication (SPA) with a lot of functionalities like buying a product, selling a product, logining ....
This application was a MVC application.

Technology used : Backbone.js, API Python, Bootstrap

Next generation 911 - Bell

This project was made to develop a proof of concept 911 call center integrated into a software showing the new functionalities of the Next Generation 911 in the lesson GLO-2004 (Project Software Engineering - University Laval)
This Software is able to simulate the travel of an Emergency Car with different intervention point where it has to go. You could modify different parameters as the speed of the vehicule, etc ...
We used mostly Java to perform this project with the Swing library.

Space Invaders

Development of a platform game with a team which consisted in designing all of the characters, the decor and the mechanic of the game.
This game in based on the same tactic of the famous game Stratego.

Technology used : ActionScript/Photoshop


Bank of Future

Social Robotics

Tablet App.

HMI : Mockups and Wireframe for a WebSite

My project was made with the University Laval.
Developed interfaces for a website as part of a university project on person-machine interface (GLO-4000)
Technology Used : Balsamiq/Photoshop/Axure

Bank of the future

The project was to think about a new bank which is more usable, more simply.
This project was more in an ergonomic way.
To realise this project, we made an analysis of needs and analysis of tasks of the users.
And we tried to display the different elements of a bank in the space we have to improve the communication.
I have made alone the video to show how we put the different things.
Technology used : Blender3D/Photoshop


The goal of the project was to link an emotion to an interaction with the robot
So we had to recognize the state of a people (smiling, sadness...) and to do the appropriate interaction.
For example, when Nao couldn't see you, he simulates the sadness.

Technology used : Python/Choregraph


One of my projects this semester is to do a tactile interface for disable kids to learn how to speak and how to write.
Their handicap is motor and cognitive. With the help of the teacher, we elaborate different exercises which are really adapted to their cognitive processes.
So, there is 3 big part of the application, the read of the words, to be able to associate a word with an image and to be able to display in the right order the letter of a word.
Of course, before the beginning of the application, the teacher has to decide the difficulty and the letter on which they will work.
With this interface, we hope that they could have a social interaction between other students.

We have also made a website with a database for the teacher to add the different pictures, and the word that she wants the pupil work on.

Technology used : C#(WPF)/Php/Javascript/MySQL

Multi Agent System

PR2 Drag&Drop

PR2 vision



We had to develop a MAS. A multi-agent system (M.A.S.) is a computerized system composed of multiple interacting intelligent agents within an environment.
Multi-agent systems can be used to solve problems that are difficult or impossible for an individual agent or a monolithic system to solve.
Intelligence may include some methodic, functional, procedural or algorithmic search, find and processing approach.

At the end, the simulation must be able to evaluate different behaviour. But we had to modelize different social behaviour like : communication, collaboration, seduction, negociation

Drag Drop

The goal of this project was to create a user interface for the robot PR2
The user could drag and drop every square on the tower.
Then, the robot will follow the pattern that the user has done on the interface
Another challenge in this project was to program the robot PR2.
This project was made at UTS (University of Technology of Sydney)

Technology used : Angular.js

Color Detection

The goal of this project was to create an algorithm which was able to detect the duplo color
We have tried different technology, but we have decided to work on OpenCV and PCL.
The robot could detect the color and the depth with the kinect.
After that, every informations are transmitted to the robot, and he catches the object he wants.
To develop the algorithm, I have used a Canny Filter to detect the edges in my images Once I have done that, I use HSV data to determine the range of the Hue, Saturation, and Value of my color Then It detects the color.
This project was made at UTS (University of Technology of Sydney)

Technology used : OpenCV, PCL

Face Detection

The goal of this project was to detect the smile or the sadness of someone with OpenCV
This project was made in C++, It was my first project using this technology.

Technology used : OpenCV, C++

Web Map

Network Analysis

Patent Mapping

MapBox connect to your photos on GoogleDrive

The goal of this project was to connect a photo database (like your public pictures on Google Drive), and to display it on a map.
This process take in count the location where you took your pictures, and display it on map with small icons.
Then, you could see all of your pictures in a map, and the locations.
For more informations, watch the video !
This project was made in Javascript.

Technology used : Js/Mapbox/GoogleDrive

Network Analysis Using D3.js

The goal of this project was to represent many publications and patent on a platform, using D3.js
This visualisation could support around 100 000 publications.
Many view are available, you could center on an organisation or just see se relation between organisation

Technology used : D3.js/Angular.js

Patent Mapping using d3.js

Geographic representation of a corpus of patent and publications
This visualisation could support around 100 000 publications.
I uses a Light technology (Web) to show and manipulate points
We have used several machine learning algorithm to be able to compute this visualisation

Technology used : Python, node.js, Javascript



IROS - Poster in a workshop - Conference IEEE
"Learning Object Affordances: a fundamental step to allow prediction, planning and tool use?"


Logo HRI

HRI - Late Breaking Report - Conference IEEE
"Natural Human Robot Interaction Using social Cues"




Postal 33710
Teuillac, France.

+33 6 99 16 79 46

Created by Hugo Romat