| Project Reference: | ITP/066/23LP |
| Project Title: | Robot Simulator for Testing and Machine Learning |
| Hosting Institution: | LSCM R&D Centre (LSCM) |
| Abstract: | With the rapid advancements in robotics technology, it has become crucial to develop efficient and reliable methods for testing and deploying robots in diverse environments. Traditional approaches involving laboratory or on-site robot testing are commonly humanintensive and time-consuming, which can impede the progress of robotics development and hinder their potential benefits. To address these challenges, we propose leveraging a 3D simulation environment for robot testing and learning. Our simulator offers a comprehensive range of functionalities, including scene creation, crowd simulation, sensor simulation, and physics simulation. By utilizing the latest multimodal language model capabilities, users can efficiently generate scenarios by inputting basic scene requirements or using reference photos. The language model also enhances the simulations by taking care of crowd behavior and physical properties, resulting in more realistic virtual environments. By harnessing these functionalities, we can create numerous scenarios and challenges for robots to navigate and interact with, all within 10 minutes. This approach significantly reduces the deployment process, enabling more efficient and effective testing before onsite trials. This development will contribute to the advancement of robotics technology, accelerating the deployment of reliable and efficient robotic solutions in diverse environments. |
| Project Coordinator: | Dr Ka Lun FAN |
| Approved Funding Amount: | HK$ 2.77M |
| Project Period: | 31 Mar 2024 - 30 Sep 2025 |
