Artificial intelligence

Artificial Intelligence Software

Description: The main objective of this project is to create a standardized, expandable, flexible and comprehensive application framework that supports the development of distributed Artificial Intelligence (AI) software based on strong modularity concepts. This framework should benefit from the latest architectures emerging from software industry experts and allow for reusability, for faster and easier software development. Ideally, a toolkit would offer flexible communication mechanisms by providing a common interface for communication while hiding the underlying protocols.

3D Synthetic Operational Environment

Description: This project seeks to investigate the causes and effects of friendly fire (a.k.a.fratricide), using various simulation scenarios executed in a synthetic environment. Two axes of investigation have been pursued as part of this project: virtual and constructive simulations.

In the constructive simulations, intelligent agents were chosen to model and simulate the soldiers’ Situation Awareness (SA) in a quick and efficient way. First, a simulation framework was developed by extending a recent version of a well-known, widely used synthetic environment: OneSAF Objective System (OOS). Then, a simulation behaviour model that details the SA and Decision-Making of a Light Armoured Vehicle (LAV) III Platoon commander was defined in accordance with a study previously made on LAV III command. By implementing this behaviour model and integrating it into the OOS environment, we were able to recreate the main factors that lead to fratricide: visual target misidentification and fear of being engaged by unidentified targets.

Intelligent Adaptative Interface

Description: The objective of this project was to develop, demonstrate and prioritize technologies that increase human info processing capabilities in all most critical applications from piloting displays, sensor operators to those involved in intelligence gathering, surveillance and reconnaissance (ISR) activities. LTI was mandated to conceive, code document and demonstrate a multi-agents architecture prototype that could be modified to handle real IAI algorithms.