Endogenous Learning by Cooperation

Context Presentation

The current digital transformation requires the creation of autonomous applications that can be adapted to complex, dynamic, heterogeneous and unpredictable environments. These systems must be equipped with proactive learning capabilities. To this end, Self-Adaptive Multi-Agent principle allows the decentralization and self-observation of the learning process. Each knowledge granule is an autonomous agent that cooperates with its neighbors to improve learning from exogenous and endogenous feedbacks. Detecting and solving concurrences, conflicts and incompetencies leads to active endogenous learning.This work on an adaptive decentralized learning mechanism will be applied on application domains such as robotics, autonomous vehicles and smart cities.


Figure 1: « Schema of the Learning

Figure 2: « Implementation Example Multi-Agent System » on an Industrial Robot »

Scientific Goals

- Design a Self-Learning System

- Lifelong and Endogenous Learning

- Genericity and Scalability


Self-Adaptive Learning, Endogenous Learning, Adaptive Multi-Agent Systems, Artificial Intelligence


bruno.dato@irit.frfrederic.migeon@irit.fr marie-pierre.gleizes@irit.fr

Automatic traffic generation with multi-agent simulation

Context Presentation

Mobility in cities is a crucial question to improve services offered to users and to decrease pollution. Traffic simulation is an interesting tool for decision-makers for city planning. Decision relevance depends on the realism of the simulation. To ensure that, we suppose that we have a map of the terrain (road topology, buildings…) and observation points that provide traffic information (throughput, direction…). They are in limited number and do not cover all the map.This internship focuses on generating car flows on the whole road network using adaptive multi-agent systems. For that, the system determines the location, speed and direction for each vehicle based on its neighborhood. The behavior of each vehicle self-adjusts to match, at the macro-level, traffic information provided by all observation points.


Figure 1: Example of traffic simulationDynameq © INRO

Scientific Goals

- Self-calibration of simulation parameters for realistic traffic flow generation

- Self-adaptation of the simulation to handle traffic dynamics


realistic traffic generation, multi-agent simulation, self-adaptation, cooperation



CLUE: Perception of the environment at the urban scale using a fleet of sensors on bicycles. Application to air pollution.

Context Presentation

Only a few fixed stations monitor air pollution at the urban scale, where it shows huge variation in space and time. However, the advent of low cost and miniaturized sensors paves the way to mobile sensor networks and crowdsensing systems. Bikes as a carrying platform seems promising: 1) distance tracks are longer than walking, 2) their embedded generator (dynamo) allows to create an autonomous energy system, 3) they do not pollute (and therefore do not distort data collection) and can both cover road network and pedestrian areas 4) human-carried measurement reinforces spatial coverage of the most frequented, hence the most important, areas. We instantiate it as CLUE: Cycle-based Laboratory of Urban Evolutions.

Auto-calibration of the sensor fleet and algorithms tolerant to erroneous measurements thanks to data density are two ways to face the low quality of the sensors (low accuracy, time drifting). Another challenge is keeping user privacy while sharing data without compromising their interpretability (for air pollution, human mobility).


Figure: CLUE embedded system

Scientific Goals

- equip a fleet of bicycles with a set of sensors

- collect information on mobility and air pollution

- merge the data of several sensors in a real environment and validate predictive models of pollutants used in aerology


Distributed sensing system, human-centered measurement tool, big data, air pollution


Christophe Bertero <christophe.bertero@laas.fr>Jean-Francois Léon <jean-francois.leon@aero.obs-mip.fr>Matthieu Roy <matthieu.roy@laas.fr>, Gilles Tredan <gilles.tredan@laas.fr>

Smart Clean Garden-Toulouse

Context Presentation

A Smart Clean Garden (SCG), is a planted filter recognised as nature based solution for water treatment of domestic water. Inspired from water quality regulation of natural rivers, a SCG shelters an enhanced biodiversity for encreased capacity of sewage in limited area of green cities. The addition of IoT as environmental sensors (moisture, NO3, Ph, conductivity,…) allows to survey and to better understand the complex system functioning inside the filter. Collected data will feed regular deterministic modeling and IA to describe the pollutant reduction process.


Figure 1 : Planted filters with 2 granulometric levels that already exist on USTH campus at Hanoi, made by Epurteck as the first SCG pilote for demonstration and logo of the project

Scientific Goals

- To demonstrate that it is possible to treat a part of domestic water of UT3 campus and producing recycled water for gardening and watering the green area of the campus

- To test the capacity building of the water purification in the planted filter by using IoT survey and environmental data modelling ?

- To identify what are the main drivers that lead to the pollutant removial in this complex system made with sediment, water , biodiversity as a micro-organisms, macro-invertebrates and plants communities, nutrients, natural organic mater and antropic molecules (fertilisers, persistant organic pollutant, medical residus, etc…


neOCampus, smart clean, garden, water, intelligent reuse, innovation,


Magali Gerino (magali.gerino@univ-tlse3.fr) ; Léo Garcia (leo.garcia@iut-tlse3.fr) ; Dan Tan Costa ( EPURTECK, dan@epurtek.fr)

Design and management of a Low Voltage DC Micro-Grid with Renewable Energy Sources and Energy Storage Systems

Context Presentation

With the environmental issues and the new ecological considerations, one of the challenge is the creation of sustainable electric grid to supply the demand. With this context, we observe the deployment of decentralized Low Voltage DC Micro-Grid (LVDC-MG) in building, with high penetration of Renewable Energy Sources (RES) and Energy Storage Systems (ESS). The aim of this PhD thesis is to contribute in this field by designing an LVDC MG in the ADREAM Building integrated PV (BiPV), at LAAS-CNRS, TOULOUSE. The main difficulties is to combined the ESS behavior and aging studies with a global system approach in order to proposed a sizing method and an energy management strategy optimized and simple to implemented for electrical research community.


Figure 1: electrical synoptic of the LVDC MG study

Scientific Goals

- Study the impacts of BiPV and DC building loads power profiles on ESS behavior and lifecycle

- Proposed a methodology, with a systemic approach, to size the PV and the ESS

- Compared multiple sizing and energy management strategy in order to design the optimal LVDC MG to supply the ADREAM BiPV lightning network

- Compared the performances of Lead acid batteries and Lithium-ions batteries in our case study


Energy Storage System, Low Voltage DC Micro Grid, Building integrated PV, Lead-acid batteries, ageing mechanisms


PhD student: mgaetani@laas.fr / Supervisors: alonsoc@laas.fr & jammes@laas.fr

Unfired earth stabilization with mineral or organic binders for sustainable and durable construction materials

Context Presentation

To mitigate the global worming related to the human activities’ greenhouse gas emissions, the so far industrial model has to be profoundly rethought. In the construction sector, ecofriendly constructions materials are gaining increasing interest as alternative to cement concrete. Unfired earth material is thus promoted thanks to its accessibility, its social, economic, and hygrothermal benefits. In the project neOCampus, our work is conducted at the Laboratory of Materials and Durability of Constructions (LMDC). The main goal is to develop unfired earth-based materials for use in potential reconstructions in the Université Paul Sabatier’s campus. Two main challenges are addressed: achieve sufficient mechanical strengths and mitigate risks related to potential water damages. To that purpose, soils from Toulouse area are stabilized with mineral binders (cement and hydrated lime) on the one hand and with organic binders (biopolymers from plants’ and animals’ byproducts) on the other hand. Unlike the modern industrial practices, a few amount of the mineral binders are used to limit their CO2 footprint. Various organic products are tested to identify the promising candidates for earth stabilization.


Figure 1: Specimens for the performances (mechanical and hygrothermal) tests

Scientific Goals

- Achieve sufficient mechanical strengths and water resistance,

- Use environmentally friendly stabilization with organic binders and mineral binders,


Construction materials, unfired earth, water resistance, mechanical strengths, ecofriendly stabilization.


Kouka Amed Jérémy OUEDRAOGO kouedrao@insa-toulouse.fr LMDCJean-Emmanuel AUBERT jean-emmanuel.aubert@univ-tlse3.fr LMDCGilles ESCADEILLAS gilles.escadeillas@univ-tlse3.fr LMDCChristelle TRIBOUT christelle.tribout@univ-tlse3.fr LMDC

Information modelling for the development of sustainable construction (MINDOC)

Context Presentation

In previous decades, environmental impact control through lifecycle analysis has become a hot topic in various fields. In some countries, such as France, the key figures for energy show that the building sector alone consumes around 45% of the energy produced each year. From this last observation emerged the idea to improve the methods hitherto employed in this field, in particular those related to the exchange of information between the various stakeholders involved throughout the lifecycle of a building. Information is particularly crucial for conducting various studies around the building; for instance, the assessment of the environmental impact of the latter. Concerning information exchange issues, the creation of open standards such as Industry Foundation Classes (IFC) or CityGML, but also semantic web technologies have been widely used to try to overcome it with some success elsewhere. Another striking issue is the heterogeneity between construction product databases. What would be particularly interesting is to know the environmental impact of a building at early phases of its lifecycle. However, there are a number of problems that still do not have solutions. This includes associating Building Information Modelling (BIM) and semantic web technologies with environmental databases to increase the flexibility needed to assess the building's environmental impact throughout its lifecycle.


Figure 1: MINDOC methodology process

Scientific Goals

- Study how information exchange is made within experts during a building lifecycle in order to figure out interoperability gaps ;

- Fill some of the encountered gaps by mean of formalization of building information.

- Combined with the formalization of environmental data on construction products, the latter will enable the introduction of product data at an early stages of the building lifecycle.


Knowledge Modeling & Semantic Reasoning - Merging Ontologies - Decision Support - Building Information Modeling (BIM) - Environmental Databases.


justine-flore.tchouanguem-djuedja@enit.fr, Bernard.Kamsu-Foguem@enit.fr, camille.magniont@iut-tarbes.fr, mkarray@enit.fr, fabanda@brookes.ac.uk.

SANDMAN: a Multi-Agent System for Anomaly Detection in Smart Buildings

Context Presentation

The number of sensors in buildings is constantly increasing, thanks to more accessible costs and the obvious interest of their use for optimized management. In this thesis we are interested in the use of data from these sensors to detect anomalies in buildings. These data, which are very numerous, can be of unknown and heterogeneous types.An anomaly is defined as unexpected and undesirable behaviour in a system and may depend on the context. In order to be able to deploy an anomaly detection system as widely as possible, it is necessary to create a decision support tool for energy experts. To address these issues, a system based on cooperative multi-agent systems implementing AMAS theory is being developed that allows anomalies to be detected by supervised learning. The anomaly detection system must take advantage of the feedback from one or more experts who label certain instances as anomalies or non-anomalies. These feedback are used for learning. The system we develop allows the addition or removal of new sensors without interrupting the detection of anomalies.



The system classifies situations


Scientific Goals

- Improve energy efficiency

- Detect anomalies in real time

- Learn continuously from the expert feedback


Multi-Agent Systems, Smart Buildings, Internet of Things, Supervised Learning, Anomaly Detection





Embedded Multi Gas Sensors for Monitoring Indoor Air Quality

Context Presentation

The measurement of indoor air quality is important for health protection against chemical and gaseous pollutants ... The indoor air can contain many pollutants such as CO, CO2, NO2 and VOCs. These pollutants exist in differents materials and products that can be used in housing (furniture, cleaners ...), but can be also comming from human activities or outside source. In this case, the detection, measurement and monitoring of these gazeuse contaminants is necessary.In view of its high performance and low cost, the innovative gas multi-sensor based on metal oxides semiconductors for analyzing and controlling indoor air quality is a good alternative to electrochemical and infrared sensors. This project is currently in progress in LAAS in collaboration with the LCC and Laplace and as part of a thesis funded by neOCampus and the Occitanie region. This thesis focuses on the characterization of multiple MOX-based gas sensors and integrates these multi-sensors in electronic card to achieve a connected object to control the indoor air quality in offices and classrooms in University Paul Sabatier in Toulouse



  Figure 1 : « MOX gas Multi-sensors»



Scientific Goals

The gas multi-sensor is a microsystem composed by four sensors on a microchip, realized to detect target gases. The scientific objective of this thesis is to characterize new nanomaterials (SnO2, CuO, ZnO, WO3 ...) designed by the LCC by using an exprimental set-up and to define an operating protocol by trying differents operationg modes.


Multi-sensors,  MOS, Indoor Air Quality, Smart building, neOCampus









Hybrid IoT: a Multi-Agent System for Persistent Data Accessibility in Smart Cities

Présentation du contexte

La réalité d'un campus intelligent ou plus généralement d'une ville intelligente passe par une observation régulière de l'environnement par des capteurs ad-hoc, afin d’agir dans l’environnement avec des dispositifs automatiques pour améliorer le bien-être des usagers. Ces capteurs permettent d’obtenir une connaissance des activités humaines et des conditions dans lesquelles ces activités sont menées, mais le déploiement d'un grand nombre de capteurs peut être coûteux. Les coûts sont principalement liés à l'installation, la maintenance et les infrastructures de capteurs dans les bâtiments existants. Pour ces raisons, l’objectif de cette thèse vise à réduire ces coûts en utilisant quotidiennement des milliers d’informations partielles et intermittentes provenant de smartphones des usagers du campus de l’Université Toulouse III Paul Sabatier. Ces traitements sont fondés sur une technologie d’Intelligence Artificielle par systèmes multi-agents coopératifs.



Figure 1 : «On utilise les informations des dispositifs intermittents et mobiles pour fournir des estimations précises»

Objectifs scientifiques

- Apprendre à partir de données brutes, imprécises et intermittentes sans feedback.

- Fournir les informations en continu, même en l’absence de données de smartphone des usagers.

- Utiliser une approche hybride de l’Internet des objets qui mixe capteurs réels et capteurs virtuels.

Mots clés

Systèmes multi-agents auto-adaptatifs, fusion de données, apprentissage, smart campus


Davide Andrea Guastella, Valérie Camps, Marie-Pierre Gleizes, {davide.guastella, camps, gleizes}@irit.fr

Back to Top