Group members: Aaron Hannon (G00347352) & Matthew Sloyan (G00348036)
Below is our demo and presentation video available on YouTube. Please click on this video image to view the video. This demo video can also be found in the "Demo Video & Presentation" folder on this repository.
This application is a virtual reality training simulation for our client. The function of the application is to reduce training costs while training ticket inspectors on the Luas in Dublin. Currently they must hire actors and close off a Luas route to train new inspectors and this project helps eliminate that. Once you launch the application on the virtual reality headset you are in a virtual train station. As soon as you hop on a train it disembarks then commencing the training session. The goal is to check everyone's ticket on board the train by conversing with all the non-player characters (NPCs) using your actual voice and they will reply to you using a text-to-speech engine (Azure). All the NPCs have different personalities so this is where the conflict resolution aspect of the project comes in. You may come across someone who may be very rude and you must coerce them into giving you their ticket or you may be fortunate to talk to someone who gives you their ticket straight away. The NPCs also have the chance to have no ticket, which is determined at random. If the NPC has no ticket the player must deal with the situation accordingly. The interactions work using an AIML chatbot that is hosted on PythonAnywhere and is connected using a flask server. Requests are made on each interaction to this chatbot and a response is predicted based on the diction we have created. Once all the NPCs are checked you may leave the train, check your score, and end the simulation. Training information such as conversation logs and session data are then uploaded to MongoDB once completed. After the purchase of a virtual reality headset there is very little cost involve and the training simulation can be replayed over and over again completely removing the need to hire actors and shutdown a Luas route for an entire day.
More information about the researched technologies and how they work together can be found in the "Disseration" folder.
git clone https://github.com/MatthewSloyan/final-year-applied-project-and-minor-dissertation.git
.