3D Detection and Quantification of Weeds in Vegetable Crops Using Drone Imaging and a Digital Twin

Innoptech is building VegForecast™, a digital twin platform for vegetable fields that combines rapid drone flyovers with computer vision and 3D reconstruction to deliver field-level insights as maps and KPIs that support scouting, intervention planning, and in-season decision making. This internship aims to deliver a complete AI module for weeds by training a detection and […]

Read More
Transformer une problématique écologique en ressource énergétique : la biométhanisation de plantes envahissantes, une approche durable pour le lac Osisko

La réhabilitation des systèmes lacustres urbains contaminés en climat froid se heurte à différentes contraintes, tant urbaines que climatiques. Le Lac Osisko est l’un d’eux : soumis à diverses pressions anthropiques, il présente des problèmes importants tels que l’eutrophisation, la contamination des eaux et des sédiments, ainsi que la prolifération d’espèces végétales envahissantes. Les stratégies […]

Read More
Development of a Store-and-Forward Navigation System with Semantic Void Detection for Autonomous Construction Monitoring

Current methods for tracking construction progress rely on manual inspections, which are slow, inconsistent, and often dangerous. This research project aims to automate this process using an agile quadruped robot (a “robot dog”) developed by Noûs Technologies. The interns will create a smart navigation system that uses the building’s digital blueprints to guide the robot, […]

Read More
ARCHAEOLOGY EXPERT SYSTEM

(1) Overview of the Research Problem Despite rapid advances in Large Language Models (LLMs), their deployment in professional knowledge-intensive sectors remains constrained by hallucinations and data gravity. LLMs frequently generate plausible but unsupported outputs when applied to specialized corpora, while organizations hold vast quantities of “dark data” embedded in unstructured PDF archives. In archaeology and […]

Read More
Self-Improving Agents with test-time compute for FDA

Current artificial intelligence (AI) agents deployed in enterprise settings make static, one-shot decisions and cannot adapt as context changes or learn from the outcomes of past actions. This research addresses two limitations of enterprise AI agents: (i) the inability to dynamically adjust reasoning effort based on the uncertainty and stakes, and (ii) the inability to […]

Read More