A/Prof Jay Katupitiya

A/Prof Jay Katupitiya Mail
School of Mechanical & Manufacturing Engineering, UNSW, Australia

The University of New South Wales

Jay is the Deputy Head of School of the School of Mechanical and Manufacturing Engineering, UNSW. His research interests include:

Agricultural Robotics

The research in this area is directed at automating large scale ground vehicles that operate on the agricultural fields and their autonomous navigation to conform to traversing specifications. They are, in general, articulated vehicle systems, in which a large tractor is providing propulsion and steering, and large implements are passively dragged behind. The control inputs to such a system are inadequate to guide the implements to follow a specified path. The research is aimed at providing the necessary control inputs at the implements to enable the precision autonomous navigation of the entire system. Research includes: i) the kinematic modelling of articulated systems taking into account the lateral and longitudinal wheels slips, ii) determination of force inputs using approximate dynamic models and the kinematic system responses, and iii) the design of controllers.

Cooperative Control of Heterogeneous Autonomous Vehicles

The research aims to achieve asynchronous cooperative control between one air vehicle and a fleet of ground vehicles. The goal is to navigate the ground vehicles to concurrently carry out a single task or to carry out a spatially distributed set of sub tasks that forms parts of a major task, under the supervision and coordination of the aerial vehicle. The fleet of ground vehicles are assumed to be distributed at different locations. In the case of a single task, the ground vehicles are assumed to have complementary capabilities. In the case of multiple tasks the ground vehicles may possess identical or complementary task execution capabilities. A practical scenario is bush fire containment. The research issues involve task identification and localisation through sensing carried out by the aerial vehicle, task scheduling, terrain mapping using the aerial vehicle, path planning for ground vehicles using terrain data and autonomous navigation of ground vehicles to their scheduled destinations. In particular, the research aims at developing a software architecture for the cooperative control of the group of vehicles and the development of methodologies for the direct navigation and control of ground vehicles under the continuous and direct command of aerial vehicles without the use of GPS data.

3-D Terrain Mapping

Methodologies are developed through fusion of multi-layered laser scans and visual images to generate a 3-D terrain map that can be used by crawling and wall climbing robots. A 7-degrees of freedom fully self contained two legged wall climbing robot has small laser range finders and miniature cameras mounted on its feet. The data gathered using the two types of sensors are fused in a special algorithm to build a map of the terrain with significant time efficiency. The 3-D surface patches identified through the procedure are then integrated to make a global 3-D map.

Mechanical Machining of Micro-lenses on Optical Fibres

It is well known that the light coupling efficiency of optical fibres can be enhanced by fabricating micro-lenses at the tips of optical fibres. This research investigates high-precision mechanical means of fabricating such lenses. Currently, the research has progressed well and purely mechanical means are used to effect efficient ductile mode machining of Silica to produce these lenses. The research project also covers mechanical means of polishing the machined lenses to further improve the surface quality so that they are suitable for optics related applications. Some example lenses machined at the tips of optical fibres are shown here.

Vectored Thrust Aerial Vehicles

A Vectored Thrust Aerial Vehicle (VTAV) is similar to a VTOL aircraft yet it is meant for hover and near hover operation. The research platform is centered around an aircraft with three thrusters. Each thruster is realized using a torque cancelling ducted fan. One of the thrusters is fixed with respect to the VTAV body and the other two thrusters can be independently vectored around a common axis. The VTAV is a lot more stable platform than a helicopter and hence has a simpler dynamic model. In addition, the dynamic models of the hover and low speed cruise are almost identical and hence cruise to low speed flight transition is easier. The research consists of dynamic model building, experimental system identification, model validation and controller design to achieve low speed, zero-roll and zero-pitch flights so that the VTAV can be used for laser based terrain mapping.

Force Controlled Ground Vehicles

This research is motivated by the need to drive an off-road vehicle very accurately in an agricultural field to follow a predefined path. The research platform is a four wheel driven four wheel steered ground vehicle that has force sensors at each wheel measuring all forces and torques. The aim is to develop a comprehensive dynamic model and to validate its parameters through experimental system identification. The control task it to vector the four drive forces and the directions of the drive forces and the side slip forces (lateral forces) so that the actual trajectory of the instantaneous centre of rotation of the vehicle will follow a desired trajectory of the instantaneous centre of rotation with constant velocity.