AI-based data fusion models for activity duration forecasting and scheduling

This project will develop an artificial intelligence-based data fusion framework for activity duration forecasting in project scheduling. Task duration estimation is an important component for project scheduling and optimization for many industries. This problem is significantly more acute for projects with a very large number of activities such as ship building and refit or when historical data is lacking. This scarcity of historical data makes activity scheduling and planning very challenging, which can lead to unplanned conflicts, insufficient resources being planned, project delays and overshooting budgets. Experts can provide estimations on task duration; however, their availability is limited, and they cannot provide estimates when there are thousands of activities to deal with. Furthermore, there is a need to better integrate the expert opinions with the limited data available from historical records or from similar equipment or projects. This will result in a decision support tool that the partner will use in scheduling and optimizing ship building and refit operations.

Faculty Supervisor:

Claver Diallo;Alireza Ghasemi

Student:

Partner:

Thales Canada Inc

Discipline:

Engineering

Sector:

Manufacturing and Construction; Advanced Manufacturing; Technology

University:

Dalhousie University

Program: