Dr. Ian Kelly

Post-Doctoral Research Fellow: Energetic Autonomy

Faculty of Computing, Engineering and Mathematical Sciences
University of the West of England, Bristol
Coldharbour Lane, Bristol BS16 1QY
UK
Phone:
Fax:


Email: ikelly@robotics.usc.edu

 

 


 

Research

After escaping from an eight-year stint at the Department of Cybernetics of the University of Reading I came to the Intelligent Autonomous Systems Lab (IASeL) to carry out research into truly autonomous robotic systems. During my self inflicted sentence of fun at Reading I completed both my first degree in Cybernetics joint with Computer Science, and a Ph.D. in Cybernetics. My doctoral research was in the fields of group robotics, shared experience learning and flocking, and basically consisted of playing with toys, err I mean building serious research robots and designing their controllers. For more serious information on my Ph.D. research and the specifications of the robots (better known as the communicating Seven Dwarfs) that I co-designed (with David Keating et al.) and constructed follow this link (communicating Dwarfs).

Whilst in the IAS Lab I designed and built a prototype robot that is able to hunt and collect slugs. The aim of this project is to construct a robotic system that is self-sufficient both in terms of computational and energetic resources. A more detailed description of this work can be found by following this link: towards true autonomy (aka the SlugBot).

I have how left the IAS lab for the more sunnier climate of the Robotics research group, University Of Southern California, 941 West 37th Place, Los Angeles, CA, 90089, USA.

 

Academic interests:

I am interested in artificial systems that take inspiration from nature. This interest is because natural systems are highly efficient and operate under many different situations. Inspiration for both my flocking algorithm and shared experience learning came from nature. The project I am currently working on is also strongly based on nature and is in essence to build an artificial predator. My main areas of academic interest are currently:

  • Robotic sensors and inter-communications between robots.
  • Physical robot system design, integration and construction.
  • Group dynamics, e.g. flocking strategies, predator-prey relationships, foraging.
  • Real time learning in real world agents.
  • Increasing the robustness and learning rate of learning agents through co-operation.
  • Incorporating ideas from nature into man-made systems, i.e. Artificial Life and Artificial Intelligence.
  • Human-robot interaction, and robot-robot interaction.


  • Publications:

     
  • I.D. Kelly, O. Holland, C. Melhusish (2000) SlugBot: A Robot Predator in the Natural World
  • I.D. Kelly, O. Holland, M. Scull and D. McFarland (1999) Artificial Autonomy in the Natural World: Building a Robot Predator
  • I.D. Kelly (1999) Towards real autonomy
  • I.D. Kelly and D.A. Keating (1998) Faster learning of control parameters through sharing experiences of autonomous mobile robots
  • I.D. Kelly and D.A. Keating (1998) Increased learning rates through the sharing of experiences of multiple autonomous mobile robot agents
  • D.A. Keating and I.D. Kelly (1998) Increasing Mobile Robot Learning Rates Through Sharing of Experiences
  • H. Hu, I.D. Kelly D.A. Keating and D. Vinagre (1998) Coordination of multiple mobile robots via communication
  • I.D. Kelly (1997) The development of shared experience learning in a group of mobile robots (PhD Thesis)
  • I.D. Kelly, D.A. Keating and K. Warwick (1997) Mutual learning by autonomous mobile robots
  • I.D. Kelly and D.A. Keating (1996) Flocking by the fusion of sonar and active infrared sensors on physical autonomous mobile robots
  • I.D. Kelly and D.A. Keating (1996) Shared experience learning on a pair of autonomous mobile robots
  • K. Warwick, I. Kelly, I. Goodhew, and D. Keating (1995) Behaviour and Learning in Completely Autonomous Mobile Robots
  • I.D. Kelly (1991) Robo-Drill : An automatic computer controlled PCB drilling machine