Mobile Intelligence for Vehicular Robots
Unmanned vehicles, including self-driving cars, have been prominently featured in news headlines lately. However, there are still numerous technical challenges to be overcome before their widespread integration into our daily lives.
Our research team focuses on improving perception and knowledge processing techniques in order to realize unmanned cars operation on public roads in real-life conditions. However, in fact, our research is not limited to car-like vehicles. We aim to develop vehicle intelligence to make various types of unmanned vehicles (e.g., cars, aerial vehicles, ships, underwater vehicles) truly autonomous.
- Enhanced Vehicle Autonomy
Unmanned vehicle technology has significantly advanced in recent years however they still need human operators for real-world applications in real-world environments. We focus on enhancing the autonomy of unmanned robotic vehicles to make them truly autonomous and applicable to real-world problems with no need for human intervention.
- Vehicle Autonomy by Incorporation of Learning, Adaptation, and Knowledge
In the navigation of an autonomous agent, intelligence in mobility includes learning aspect. Advancing from conventional robot navigation that understanding map statistically, a new robotic mapping needs to be smart by learning and adapting the map in order to cope with both long-term and short-term changes.
- Active SLAM and Active Perception
For a fully autonomous navigation a robot needs to handle confusion. When the robot thinks it is lost, the reaction should follow to increase accuracy in localization, as human’s behavior of looking around and checking maps. This active reaction and control to correct navigation error incorporated navigation enable a robotic agent to navigate the environment with minimum human intervention.
- Cooperation between Multiple Unmanned Vehicles
Unmanned vehicle technology has matured to a point that a single vehicle can be used to perform a simple mission. Current mission capabilities and performance can be greatly increased by using multiple vehicles simultaneously. Our research aims to develop new techniques and algorithms for cooperative/collaborative operations with multiple unmanned vehicles.
- Guidance, navigation, and control
- Perception and sensor fusion
- Motion planning
- Mission/task allocation