Magic happens in UTS lab
- UTS PhD student Rony Novianto is the sole Australian recipient this year of a prestigious IBM PhD Fellowship
- The intensely competitive worldwide fellowship program honours exceptional PhD students working on solving problems that are important to IBM and fundamental to innovation in many academic disciplines and areas of study, in Rony's case his groundbreaking work on cognitive software architecture
IBM PhD Fellowships are hard to come by, as UTS PhD student Rony Novianto will attest. Rony is the sole Australian recipient of a 2011 award, thanks to his groundbreaking work on cognitive software architecture.
Rony is a PhD student in the Centre for Quantum Computation and Intelligent Systems Magic Lab, more formally known as the Innovation and Enterprise Research Lab. He is developing a novel cognitive software architecture that can be applied to self-directing systems, such as software agents and autonomous robots.
The architecture is called Attentive and Self-Modifying (ASMO). When it's applied to self-directing systems, ASMO orchestrates information from a range of sensory data to solve complex problems.
Unlike most cognitive software architecture, ASMO has been designed based on the integration of practicality and functions of the human mind. It allows software agents and robotic systems to respond to new and changing conditions and to deal with unstructured problems in complex and dynamic environments. The architecture could lead to the development of a wide range of innovative applications such as IBM Smarter Planet systems, advanced rescue robots used in disaster recovery work, or robots that can be used at home.
"Today's robots are very successful and used extensively in a range of structured industry settings, but they are limited in other areas such as homes," Rony said. "One of my research goals is to transform robotic systems to be used more widely in these unstructured areas so they are more accessible to mainstream users."
Current autonomous systems are limited to pre-defined situations – that is, software designers anticipate user inputs and program the corresponding responses into their systems. However, this is ineffective when a system is functioning in an environment where actions and behaviours for all possible conditions can't be specified – for example, in an unstructured human environment where choice behaviours are impossible to predict.
Rony has previously used ASMO on the UTS/University of Science and Technology of China robot soccer team, which competed at last year's RoboCup. While many other teams still use traditional finite-state machines, Rony believes that ASMO gives the UTS team an edge. ASMO is easier to use, more flexible, and provides faster development which allows the soccer robots to integrate various models developed by different developers.
Glenn Wightwick, Director of IBM Research and Development - Australia, said that the intensely competitive worldwide fellowship programme honours exceptional PhD students working on solving problems that are important to IBM and fundamental to innovation in many academic disciplines and areas of study.
"We have received applications from outstanding candidates at universities all over the world in many areas including computer science and engineering; electrical and mechanical engineering; physical sciences; mathematical sciences; business sciences; and service science, management and engineering," Mr Wightwick said.
Rony is the second UTS QCIS and Magic Lab student to win an IBM PhD scholarship, following on from last year's winner, Xun Wang. Award winners receive access to a range of internships and activities to further their research, and have the chance to undertake internships with mentors who are leading the way into the future of computing.
"IBM is the leader of innovation in artificial intelligence (AI), so this will be an amazing experience to work with many AI experts," Rony said. "It's a fantastic opportunity to build my future career, to help build a smarter planet and to contribute to society."
by Claire Thompson
(Media enquiries) Terry Clinton (+61 2 9514 1623)