The Automated Driving and Human-Machine System (AutoMan) Group @ NTU has a wide range of research activities covering design, control and optimization of human-centric robotics, with particular focuses on safe, smart and sustainable vehicles, smooth human-machine interaction, as well as the science of cyber-physical systems. AutoMan group needs a cost-effective car-like robot platform to conduct, validate and verify their research.
The robot consists of a list of devices that are similar to those mounted on real autonomous driving car, e.g., 3D lidar, with lidar-based SLAM and navigation. To overcome the power consumption issues, the robot equips with low-power Nvidia Jetson NX board for navigation and other machine learning algorithms. The whole system is ROS compatible.
Understand the surrounding environment and human behaviour is one of the key active human-robot interaction research topics. The robot installed both RGB and RGBD cameras in the front and the back. To achieve better collision avoidance capability, 8 ultrasonic sensors are mounted.
With this low-cost research platform, researchers are able to quickly develop, deploy and test their research algorithms inside the lab or the campus. Once verified, the algorithms can be migrated to a real autonomous driving car for further validation and verification.