RESEARCH STUDENTSHIPS AVAILABLE

Shortlisting will begin on 5th September 2005

The PhD studentships offer the opportunity to work within the Adaptive Systems Research Group at University of Hertfordshire,a proactive and dynamic research team with an excellent internationalresearch profile. The group was founded and is co-organized by Prof.Kerstin Dautenhahn and Prof. Chrystopher Nehaniv. Other core facultymembers of the group include Dr. René te Boekhorst, Dr. Lola Cañamero,and Dr. Daniel Polani.  Current projects within the AdaptiveSystems Research Group are funded by FP6-IST, EPSRC, the Wellcome Trustand the British Academy. We are currently involved in the following FP6European projects: Euron-II, Humaine (both European Networks ofExcellence), Cogniron and RobotCub (Integrated Projects). We hosted theAISB'05 convention "Social Intelligence and Interaction in Biologicaland Artificial Systems" that attracted an international audience of 300participants. Research in the group is highly interdisciplinary andstrongly biologically inspired, but also has a strong theoreticalfoundation.

Adaptive Systems are computational, software, robotic, or biologicalsystems that are able to deal with and "survive" in a dynamicallychanging environment. We pursue a bottom-up approach to ArtificialIntelligence that emphasizes the embodied and situated nature ofbiological or artificial systems that have evolved and are adapted to aparticular environmental context. The Adaptive Systems Research Grouphas excellent research facilities for research staff, includingnumerous robotic platforms, covering the spectrum from miniatureKhepera robots to human-sized robots (PeopleBots), as well as humanoidrobots developed in our group. We are dedicated to excellence inresearch, and, while providing a collaborative and supportive workingenvironment, expect PhD students to show great enthusiasm anddetermination for their work. Candidates need to provide evidence ofexcellent research potential that can lead to significant contributionsto knowledge as part of a PhD thesis.

Applicants are required to have a very strong first degree or(preferably) a specialist Master's degree in Computer Science,Cybernetics or a related area relevant to the research theme.Candidates should be highly motivated and have an excellent backgroundin Artificial Life, Artificial Intelligence and (where relevant to theresearch theme) Behaviour-Based Robotics. Very strong programmingskills are an essential requirement, as well as general interest ininterdisciplinary research and willingness to collaborate withresearchers from other disciplines.

This year, the Adaptive Systems Research Group advertises PhDstudentships associated to five proposed themes that can lay thefoundation for a concrete PhD project. Applicants interested inparticular themes should use the contact email provided for informaldiscussions regarding suitability for the research themes. All formalapplicants must be made through the Research Office, see contactinformation below.

1. Behavioural Analysis and Motivational Modeling of Autonomous Agents (contact: R.teBoekhorst AT herts.ac.uk)
This research theme addresses the following two subthemes that couldform part of PhD research: a) the emergence of "social" organisationcharacterised by role diversification and division of labour inself-organised "robot societies" where the behaviour of the agents isbased on direct interaction instead of global knowledge (situatedness),distributed organization instead of central control and form/physicsinstead of purely algorithmic computation (embodiment), candidatesshould possess a good background in swarm intelligence and collectiverobotics, b) formal models of motivational dynamics and theirimplementation (in robots) resulting in a robotics architecture basedon ethological models and studied by means of ethological methods. Arecently developed model to explain changes in motivational intensitybased on simple considerations about energy allocation, negativefeedback (like the inhibitive workings of "consummatory behaviour"),positive feedback (fear brings more fear) and the impact of externalstimuli can be used as a starting point.  Candidates shouldpossess excellent skills in mathematics and be motivated to studyethological theories about motivation.
 
2. Evolution of Sensors, Constructive Biology and the Perception-Action Loop (contact: D.Polani AT herts.ac.uk)
A core activity of any viable biological and artificial adaptive agentconsists of the acquisition of information from its environment, theprocessing of this information and the subsequent selection of actions.Based on recent progress in the use of information-theoretic,statistical, and automata theoretic methods the group has identified awide spectrum of fundamental and powerful first principles governingthe structuring of perception-action loops in agents. This workcontributes to our understanding of fundamental phenomena in theemergence of intelligence and life and leads to real-world applicationsincluding robotics, evolutionary biology, evolution of communication,control and sensor fusion, and the deployment of self-configuring and-adapting sensors in practical applications. Applicants are expected tomake a significant contribution to the group's research activities insensor evolution. Specific PhD projects can include more theoreticallyoriented work as well as using robotic platforms as scientifictest-beds.
 
3. Social Robotics and Human-Robot Interaction (contact: K.Dautenhahn AT herts.ac.uk)
As part of the European projects Cogniron and Robotcub we investigate aspects of social interaction and communication in human-robotinteraction, including: computational mechanisms for the development ofsocial interaction, communication skills and interaction histories in robots; social learning and imitation in the context of a robot learningnew skills from a human teacher; and computational mechanisms ofadapting a robot's behaviour styles towards preferences of users(prerequisite for personalized service robots). The work will bestudied using robotic platforms, this might require interfacing newsensors or actuators or slight modifications to the hardware. In aproject on assistive robotics (www.aurora-project.com)we study how robots can contribute to the education and therapy ofchildren with autism, addressing issues of imitation, joint attention,and play skills in interactions of children with robots.
 
4. Biologically-Inspired Affect-Based Regulatory Systems for Robots (contact: L.Canamero AT herts.ac.uk)
Emotion is not regarded anymore as an undesirable consequence of ourembodiment that must be neglected, but as an integral component ofintelligent behaviour that offers a rich potential for the design ofartificial adaptive systems and for enhancing our interactions withthem. Within the general framework of biologically-inspired AI androbotics and affective computing, research projects can include a)investigating the adaptive value of different affective phenomena tomodulate and control the behaviour and interactions of autonomousrobots in physical and social environments, involving implementationand experimental testing and analysis of suitable mechanisms andarchitectures underlying such affective phenomena, b) studying therelationships between emotional and cognitive development, and inparticular how affective mechanisms influence the development ofcognitive and behavioural mechanisms and skills, c) research intodifferent aspects of emotions and their expression in human-robotinteraction involving e.g. an expressive robotic head that wouldinteract with humans. Strong interdisciplinary research interests(including neuroscience, robotics, psychology) are required.

5. Developmental Genetic Regulatory Networks (DGRNs) (contact: C.L.Nehaniv AT herts.ac.uk)
We are studying the evolvability of multicellular genetic regulatorynetworks and assessing their potential to generate control andinteractive dynamics for various applications, including the modelingof real organisms, robot control, and continual adaptive responsivenessvia phenotypic plasticity (e.g. in self-maintenance, growth, androbustness).  Of particular interest are the computationalrealization of differentiated multicellularity for producing dynamicadaptation, division of labour via duplication and divergencemechanisms, dealing with the above-mentioned problems, and/orcomputational morphogenesis. Also of interest is the possibility ofusing DGRNs as a model for a non-von-Neumann computational paradigmthat can cope with and take advantage of changing environmental andinteractional constraints. Candidates require a background thatincludes foundations in relevant biological aspects (evolution,genetics, development) and/or their analogues in artificial life.

Successful candidates may be eligible for a research studentship awardfrom the University in some of these areas (equivalent to about £9000per annum bursary plus the payment of the standard UK student fees).Self-funded students might also consider to pursue other researchtopics that senior academic members of the Adaptive Systems researchgroup are active in, please consult http://adapsys.feis.herts.ac.uk/ for more information.

For further information on the individual topics advertised and an application form, please contact:

Mrs Lorraine Nicholls,
Research Student Administrator,
STRI,
Faculty of Engineering and Information Sciences,
University of Hertfordshire,
College Lane,
Hatfield, Herts, AL10 9AB
United Kingdom