Optimization of the Supply Chain Process for Aircraft Film Solutions- QC-362

Desired discipline(s): Computer science, Mathematical Sciences, Operations research
Company: Libellule Monde Inc. (LBM)
Project Length: Flexible
Preferred start date: As soon as possible.
Language requirement: Flexible
Location(s): Saint-Jérôme, QC, Canada
No. of positions: flexible
Preferred institutions: McGill University, Polytechnique Montréal, Université de Montréal

Search across Mitacs’ international networks - check this box if you’d also like to receive profiles of researchers based outside of Canada: 

About the company: 

LibelluleMonde (LBM) analyzes, designs, produces, installs and removes aircraft decals, stencils, signage, aesthetic and (permanent/temporary) protective coatings for all types of aircraft. 

These are used for a variety of applications (industrial, aesthetic, protective and branding/embellishment) for all types of surfaces in specialized and regulated sectors. 

We have production, distribution and CRM data collected over 25 years as well as privileged access to confidential data by Air Canada. These collective datasets reflect time lost related to purchasing, repairs, leasing, waste, number of POs, pricing, manhours, shipping, etc.

Please describe the project.: 

The project focuses on the development of AI-based solutions to improve production, logistics and supply chain processes (Machine Learning, Data Science, NLP, Computer Vision).

We identified 5 different themes, which can lead to projects on their own or addressed as a whole, to be discussed with participants:

1) Reduction & Replacement of Recurring & Repetitive Tasks:

The main objective is to reduce internal manual tasks and errors, from the management, production and shipping teams.

We are considering NLP and ML to be implemented into our product line and into our dynamic system.

2) Kit Assembly Optimization:

The goal is to have a 3D model of the aircraft segregated into several sets, each representing a part of the aircraft. The 3D model is textured using real images so it allows the platform to process all computations using an up-to-date version of the aircraft.

We are looking for expertise in Machine Learning, 3D algorithms development as well as a time series analysis.

3) Comparison & Compliance through the use of Artificial Intelligence (AI)

This subproject addresses the inclusion of 3rd party companies under the LBM regulatory authority, where 3rd party processes will be scrutinized through an AI exercise, measuring them against LBM certifications and processes.

The challenge is to build a multipurpose preprocessing framework to transform our raw text into a form that is ready for computation and modelling. The dictionary of objects must be relevant and industry-specific.

4) Inventory Optimization:

LBM wants to develop an AI Inventory Optimization Model, coupled with consignment and the build-up of LBM kit assemblies and RFID implementation on all parts.

5) Customer Service Assistance:

We aim at building automatic translation algorithms in 25 languages using NLP in-house code based on CNN & LSTM algorithms to ensure the high-quality fidelity of translations specific to this highly regulated sector where terminology must be mastered and understood to ensure compliance.

Furthermore, AI integration will allow for the planning and preparation of customer requirements through a faster and safer process, and automated customer services will provide clients with a menu of multiple options to choose from, as a recommendation agent.

Required expertise/skills: 

Doctoral or Master's student.

Specific software or skills:

  • Very good knowledge of Python, PyTorch, TensorFlow, Scikit-learn or XGBoost;
  • Knowledge of the Cloud, Azure being a strong asset;
  • Experience in Artificial Intelligence and/or Big Data projects;
  • Good mathematical knowledge of machine learning
  • Experience with Data Stream systems: Storm, Sparks-Streaming