Since 2016, the Sophia Antipolis center has implemented the Mutual Action for Technological Development (or AMDT).
The AMDT is a major Technological Development Action whose purpose is to develop high-level softwares for Inria research teams while sharing code and good practices. The developments are led by a team of developers of 7 to 8 engineers with complementary profiles. Agile methods allow them to develop functional deliverables in a short time (one-month cycle) in an iterative way. The focus is on the pooling of production and know-how, but also on the ability to produce demonstrative softwares and closer to the transfer to industry and scientific communities.
Achievements
The AMDT team is involved in the development of several software and scientific platforms.
Precis
Precis |
SDM Modeler
SDM Modeler |
Multi-domain platform
Multi-domain Platform |
Control Toolbox (ct)
This project, led by the team McTao in collaboration with the CAGE team from Inria Paris and the CNRS (Toulouse), aims to establish a collaborative platform that aggregates reference tools to solve optimal control problems. The platform links two essential building blocks: [bocop] (https://www.bocop.org/) and [hampath] (http://www.hampath.org/). The idea is to use Bocop to initialize HamPath to determine the optimal control structure and active state constraints, if any. The main technologies used are C ++
, Fortran
,dtk / Qt
.
Here is examples of notebooks from the ct (control toolbox) project using nutopy package and others.
- Optimal Control Problems (OCP)
- Problèmes Valeurs Initiales (IVP)
Kepler (minimum time orbite transfer) | Kepler (with conjugate point test) | SIR model |
Gnomon
Gnomon is a platform supported by the team Mosaic of the Inria center of Grenoble and dedicated to the simulation and analysis of biological systems in development (morphogenesis).
Gnomon is based on a platform / plugin architecture. The core of the platform written in C ++ is based on dtk / Qt
and the plugins are written in either C++ or directly in python. The C++/python wrapping is based on SWIG
and SIP
, the visualization part is based on dtk/VTK
and the plotting tools come from matplotlib
.
Gnomon Version 1.x.x | Gnomon Version 2.x.x |
SW-Platform
SW2D (Shallow Water 2D) is a platform written in C ++ supported by the LEMON team and dedicated to modeling shallow water flows with additional features such as porosity modeling (upscaling), passive transport, multilayer models.
From a technological point of view, this platform makes use of a number of libraries including Qt
, dtk
, VTK
and xtensor
.
The following video illustrates the work done by the AMDT team and the Montpellier LEMON team after two 15-day sprints.
More info on SW2D platform page.
Lac Taihu with standard SW équations |
Macular
Macular is a platform supported by the Biovision team and dedicated to large-scale simulations of the retina under normal and pathological conditions.
This C ++ platform is based on dtk
,Qt
, VTK
,xtensor
and GSL
for ODE solvers.
Graph Generator | Retinal Waves | Prosthesis simulation |
Bolis2
The purpose of the Bolis2 ADT is to develop the Fs3d (or FindSources3D) software dedicated to source localization by solving inverse electroencephalography (EEG) problems. From spot measurements of electrical potential, obtained numerically or taken by electrodes on the scalp, Fs3d estimates point dipole current sources in the brain. This project involves the teams FACTAS, ATHENA of Inria Sophia Antipolis Méditerranée as well as the Center of Applied Mathematics (CMA) of the Ecole des Mines de Paris.
From a technical point of view, Fs3d is based on a Matlab
computing core wrapped within a C ++ framework usingdtk
, Qt
andVTK
.
Fs3d demo |
Odin+
Odin is a platform for advanced monitoring of bioreactors by non-linear control algorithms with applications for methanation and microalgae production. It is developed for the team Biocore.
Odin is based on an n-tier architecture where each element communicates via an MQTT broker. To ensure robustness and fault tolerance, the acquisition, calibration and monitoring components are written in Erlang. The are written in Erlang. The user interface is in the form of a WEB application written in javascript and using nodejs
. The control algorithms are written in python.
BCI-Browser
This project led by the team Athena aims to introduce BCI (Brain Computer Interface) type interactions into existing applications (such as a typical Web Browser).
The architecture is of client / server type written in C ++ and using Qt
,OpenVibe
and EEG acquisition drivers.
FlowNext
This project aims to equip the Igloo software suite with a GUI allowing its interactive use and the interoperability of its tools. Igloo is developed by the Acumes team and integrates standards in geometry (NURBS) and simulation (FE, DG) under a single paradigm. The goal is to provide more efficient interactions between geometry and analysis and to converge towards a new paradigm for industrial design.
The GUI is written using Qt
and the visualization is based onOpenGL
.
Igloo demo |
MGDA
The Web Portal MGDA provides tools for differentially and / or multi-point (unstationary) multi-discplinary optimization.
It is written in javascript and uses nodejs
. The underlying library is written in Fortran with a C ++ wrapper. Users can upload their data and retrieve the results after running the code.
Patient Monitoring
This project is conducted with the team Stars and aims to develop a platform dedicated to the monitoring of patients with behavioral disorders.
The platform integrates modules for acquisition, detection of people and events. It relies on a platform / plugin architecture from dtk
and makes use ofPython
, Qt
,OpenCV
and Machine Learning algorithms.
Sup demo |
WindPos
This project aims at equipping the SDM solver of the Tosca team with a pre and post treatment tool for large scale forcing and topography information.
SDM-Modeler is written in C ++ and uses dtk
,Qt
and VTK
.
Authors
Graduated in CFD, Thibaud Kloczko is the head of a research engineer team dedicated to scientific software development and experimentation at INRIA. As a software engineer from 2011 to 2019, he was first involved in the development of the meta platform dtk that aims at speeding up life cycle of business codes into research teams and at sharing software components between teams from different scientific fields (such as medical and biological imaging, numerical simulation, geometry, linear algebra, computational neurology). He also helped set up agility for the development of scientific software at the Sophia Antipolis Inria Center.