Prédiction et optimisation de la main-d'oeuvre maritime au port de Montréal
Bibliographic record
Abstract
RÉSUMÉ: Au port de Montréal, une nouvelle solution est recherchée pour optimiser la charge de travail quotidienne liée au trafic maritime ainsi qu’au volume et au type de marchandises qui y transitent. Le projet Galileo, financé par le consortium ScaleAI [1], a pour objectif d’offrir des estimations plus précises pour les besoins de main-d’oeuvre liés au chargement et déchargement des navires, complétant ainsi les ambitions du Port de Montréal en fournissant une visibilité inédite sur la chaîne d’approvisionnement. C’est dans le cadre de cette initiative que ce mémoire se concentre particulièrement sur le problème que constituent les allocations inefficaces de ressources humaines. Ces dernières sont intrinsèquement liées à la difficulté pour un opérateur humain d’évaluer avec précision les besoins en main-d’oeuvre au-delà d’un horizon à court terme, la charge de travail au port fluctuant fortement en fonction de la densité du trafic, de la météo, de la nature des marchandises et des disponibilités du personnel [2]. Ce mémoire contribue donc aux efforts conjoints d’Airudi Inc. et de l’Association des Employeurs Maritimes (AEM) dans la conception d’un outil qui permet de prédire avec précision la main-d’oeuvre requise jusqu’à 21 jours à l’avance, améliorant ainsi les performances et la fluidité opérationnelle au Port de Montréal [3]. La méthode présentée ici comprend donc une stratégie originale et innovante alliant apprentissage automatique et modèles d’optimisation sous contraintes. L’association de ces techniques permet à la fois de produire des indicateurs statistiques prévisionnels tangibles et conformes aux disponibilités du personnel ouvrier. ABSTRACT: At the Port of Montreal, a new solution is sought to optimize the daily workload related to maritime traffic, as well as the volume and type of goods passing through. The Galileo project, funded by the ScaleAI consortium [1], aims to provide more precise estimates of manpower needs for loading and unloading ships, thus complementing the ambitions of the Port of Montreal by providing unprecedented visibility into the supply chain. It is within the framework of this initiative that this thesis focuses particularly on the problem of inefficient human resource allocations. These are inherently linked to the difficulty for a human operator to accurately assess labor needs beyond a short-term horizon, with port workload fluctuating greatly depending on traffic density, weather, nature of goods, and staff availability [2]. This thesis therefore contributes to the joint efforts of Airudi Inc. and the Maritime Employers Association (MEA) in designing a tool that accurately predicts labor requirements up to 21 days in advance, thereby improving performance and operational fluidity at the Port of Montreal [3]. The method presented here thus encompasses an original and innovative strategy combining machine learning and constraint-based optimization models. The association of these techniques allows for the production of tangible predictive statistical indicators that are in line with the availability of labor.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".