01 / RESEARCH
Research interests
Visual navigation, mapping, and multi-sensor state estimation for spacecraft and robots operating with limited prior knowledge of their surroundings.
Navigation & mapping
PRIMARY RESEARCH
SLAM for small-body exploration
- Joint estimation of spacecraft motion and environmental geometry from visual observations.
- Integration of factor-graph smoothing, recursive filtering, and orbital dynamics for proximity navigation.
- Metric-scale observability and uncertainty under target rotation, variable illumination, and uncertain dynamics.
- Computational efficiency and estimation consistency within onboard processing constraints.
Perception
VISION & LEARNING
Visual measurements for navigation
- Feature tracking and multi-view geometry for motion estimation and scene reconstruction.
- Photometric modelling of visual measurements under variable illumination and viewing geometry.
- Learning-based correction of image-derived navigation measurements, including target centroid estimates.
- Event-based vision for odometry and localization.
Multi-sensor estimation
ROBOTICS
State estimation for robotic platforms
- Tightly coupled fusion of visual, inertial, joint-state, and contact measurements.
- Motion estimation for humanoids with free-floating or underactuated bases.
- Complementary geometric, kinematic, and contact constraints for localization.
- Assessment of sensor contributions, estimation accuracy, and uncertainty.
Simulation & validation
SIMULATION & VALIDATION
Modelling and evaluation of navigation systems
- Spacecraft dynamics models with gravity, solar-radiation pressure, and manoeuvres.
- Synthetic sensor measurements for navigation-system development and validation.
- Spectral and radiometric rendering under variable illumination, surface properties, and viewing geometry.
- Monte Carlo assessment of estimation errors, uncertainty, and computational cost.
