Exploring Autonomous Navigation and Decision Making with Mobile Robots in Simulated Escape Rooms Using Deep Reinforcement Learning
This repository is part of a comprehensive study conducted by researchers at Northeastern University to apply Deep Reinforcement Learning (DRL) techniques—specifically Deep Deterministic Policy Gradient (DDPG) and Twin Delayed Deep Deterministic Policy Gradient (TD3)—to enhance the autonomous navigation and decision-making capabilities of mobile robots within simulated escape room environments equipped with differential drives. Our work demonstrates significant advancements in the robots' problem-solving abilities and adaptability, which are critical for real-world applications like disaster response and automated warehousing.
This project investigates the performance of advanced DRL algorithms in scenarios featuring continuous action spaces, enabling robots to efficiently reach specific positional targets using kinematic motion models. The insights gained underscore the transformative potential of DRL in developing sophisticated navigational strategies in complex, dynamic environments.
- To enhance the cognitive and navigational abilities of autonomous robots.
- To demonstrate the efficacy of DRL algorithms in complex scenario navigation.
- To explore applications in critical areas such as disaster response and search and rescue operations.
- Python 3.x
- PyTorch
- OpenAI Gym
- NumPy
- Matplotlib
Clone the repository: --- coming soon ...