Développement des outils en vue d’une preuve de concept de casque intelligent pour pompier

Le sujet de la recherche concerne l’évaluation de l’ergonomie de l’interface graphique et la mesure de la charge cognitive ainsi que l’évaluation des cas d’utilisation du casque intelligent pour pompier en situation d’intervention en vue de proposer des ajustements ou améliorations ou encore de nouveaux cas d’utilisation

Enabling safe multicore utilization for ARINC-653 based systems

Airborne systems (i.e. control software of airplanes) are a special type of system in which a set of rules (i.e. ARINC-653 and DO-178) should be respected, due to their criticality. Currently, most of these systems execute on architectures with only one core (i.e. processor). In architectures with multiple cores, some components of the architecture are disabled, in order to prevent interference, not acceptable following aforementioned rules.

Hierarchical graph kernels for classification of molecules

The central problem of pharmaceutical research is to understand the effect of a certain molecule on human or animal biology. Many machine learning models have been developed in recent years and show great efficiency and accuracy for this kind of predictions. However, a problem common to many of the best algorithms is that they require huge amounts of data to be sure the models are well trained and make accurate predictions. This is a serious issue when such data volume is not available.

Development of prediction models in smart buildings using data driven approaches through Artificial Intelligence

BrainBox AI is working on the mandate to provide better solutions for optimized control of HVAC systems modeled through data using AI. The Intern will continue our research at BrainBox AI and to continue to develop data driven approaches for an optimized operation of HVAC systems in design and implementation of scalable AI frameworks for AI needs in the ongoing projects at BrainBox AI. The Intern will apply AI (including ML/DL/RL), data mining and statistical approaches for the creation of scalable predictive models.

Évaluation des écarts dans les résultats d’analyse du cycle de vie des bâtiments

L’utilisation des outils d’analyse du cycle de vie permet de réduire considérablement l’impact environnemental des bâtiments. À l’aide de ces résultats, les concepteurs peuvent améliorer la performance environnementale de leurs bâtiments. Toutefois, l’utilisation de ces outils dans le domaine est encore très peu répandue. Afin de faciliter leur adoption, plusieurs fournisseurs de logiciels proposent des alternatives pour les concepteurs et analystes.

Design of a Force-Feedback Manipulandum with Integrated Sensors

Haply Robotics is developing a haptic device that produces forces on the hand while interacting with medical simulations in virtual reality. As part of this larger project, we will design and evaluate an improved manipulandum – the stylus of the device that is held in the hand – with integrated sensors capable of measuring the grip force applied by the hand and the location of the manipulandum.

Bond Pricing AI Improvement

The fixed-income market consists of government and corporate bonds and other debt instruments which are used to finance operations and capital investments. The bond market remains heavily reliant on exchanges of information between counterparties and as a result information on prices is decentralized and market participants operate with different levels of information. The objective of this research project is to create improved Artificial Intelligence models which will allow market participants to better manage trading activities, manage risk, or make portfolio funding allocations.

Séparation magnétique appliquée à l’extraction de la magnétite dans les résidus de bauxite calcinés

Le but de cette étude est de réaliser une séparation magnétique de la magnétite (oxyde de fer) contenue dans le résidu de bauxite calciné, dans le but de le valoriser. L’étude se déroule en plusieurs volets : la caractérisation du résidu de bauxite, phase expérimentale de séparation et le développement d’un modèle mathématique. La caractérisation permet de connaitre entre autres la distribution granulométrique et les espèces chimiques. La phase expérimentale permet d’identifier les paramètres d’influence de la séparation. Le modèle mathématique permettra de réaliser des simulations.

Tackling metabolic diseases using high throughput mass spectrometry coupled to artificial intelligence algorithms

This project will provide an artificial intelligence-based tool to predict biomarkers associated to metabolomic imbalances in multiple cell types and disease states. Thanks to the large amount of mass spectrometry data on which state-of-the-art machine learning algorithms will be trained, the software solution will achieve high accuracy, clinical-grade, predictions. Ultimately, the software will provide possible targets for small molecule therapy.

Unsupervised Learning of 3D Scenes from Images using a View-based Representation

We’d like to address the issue of 3D reconstruction from 2D images. This means developing a machine learning algorithm that can take a regular photo as an input and generate a full 3-dimensional reconstruction of the contents of the photo. Such technology can be used creatively or to help the coming generation of robots better understand their surroundings.