Embedded Files
DEMO
DEMO
INTRODUCTION
INTRODUCTION
This work proposes using Virtual Reality (VR) to conduct effective industrial worker training. Reinforcement Learning - Belief Desire Intention (RL-BDI) agents are integrated to aid user decision-making. The realistic VR environment facilitates collaborative efforts and productivity optimization. RL-BDI agents have human-like capabilities to enhance learning. A 'Gradual Training' approach is proposed to train agents faster. Significant findings include improved agent accuracy and success rate in performing tasks. This methodology ensures enhanced safety, cost-effectiveness and decision-making through immersive VR training.Â
This work proposes using Virtual Reality (VR) to conduct effective industrial worker training. Reinforcement Learning - Belief Desire Intention (RL-BDI) agents are integrated to aid user decision-making. The realistic VR environment facilitates collaborative efforts and productivity optimization. RL-BDI agents have human-like capabilities to enhance learning. A 'Gradual Training' approach is proposed to train agents faster. Significant findings include improved agent accuracy and success rate in performing tasks. This methodology ensures enhanced safety, cost-effectiveness and decision-making through immersive VR training.Â
KEY FEATURES
KEY FEATURES
Unity ML-Agents: Trained Reinforcement Learning - Belief Desire Intention (RL-BDI) agents to improve user decision making during training.
XR Interaction Toolkit: Used XR interaction for amplified immersive experience.
Gradual Training: The Gradual Training approach incrementally increases the complexity of tasks over time, starting with a simplified environment to facilitate efficient learning and gradually introducing harder environments to decrease the learning time for the agent.
SKILLS
SKILLS
Language: C#
Technology: Virtual Reality (VR) Integration, Reinforcement Learning (RL), Unity 3D, Head Mounted Display (HMD), Artificial Intelligence (AI), Serious Gaming Development
PROJECT LINK
PROJECT LINK
Page updated
Google Sites
Report abuse