AI development for misty robot system
The efficiency of front desk operations has been shown to improve with the adoption of AI-enhanced service robots. However, the adoption of these robots for front desk operations in the events industry has been limited, possibly due to cost constraints or concerns about user acceptance and integrati...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
Nanyang Technological University
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/181454 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | The efficiency of front desk operations has been shown to improve with the adoption of AI-enhanced service robots. However, the adoption of these robots for front desk operations in the events industry has been limited, possibly due to cost constraints or concerns about user acceptance and integration challenges.
This report documents the deployment of the Misty Robot for guest registration and question-answering purposes at events. This project employs Misty’s built-in capabilities, such as object recognition, QR code scanning, text-to-speech, speech capture, and expressive facial emotions, to create an interactive registration experience. The project involves four main components: the Misty Robot at the front desk, a Next.js Web application for interfacing with Misty and keeping track of guests’ registration status, a MongoDB server for backend storage, and a FastAPI server with a Retrieval-Augmented Generation (RAG) process to generate relevant responses to guest queries using a Large Language Model (LLM).
This project aims to develop a framework for the Misty Robot to enhance guest registration efficiency, improve guest satisfaction, and create a more engaging experience at events and exhibitions. Additionally, it aims to utilise the Misty Robot to actively engage guests by providing accurate and relevant responses to their event-related queries through a RAG-based LLM, thereby significantly enhancing their overall event experience and satisfaction. |
---|