The Centre for Robotics & Intelligent Systems

The Centre for Robotics and Intelligent Systems houses a multidisciplinary group with interests in both static and mobile robotics and their constituent technologies. Biologically inspired control and behaviour is a common theme in much of the group's work. Applications include the development and study of miniature robot 'bugs' and myoelectric teleoperation. Dynamic vision sensing is another common theme in many projects. These include applications as diverse as the identification of marine plankton and autonomous robot navigation. Intelligent, multi-agent behaviour, as applied to Mirosot robot football, provides an interesting focus for group activities. The group has strong national and international links with both industry and research institutes. It is represented on the board of the British Automation and Robot Association, the IEE and EU funded projects. Members are listed below following a brief summary of some current projects.

Staff List

The Plymouth Hand Project

Contact: Mr. Paul Robinson
Tel: +44 (0) 1752 232595
p.robinson@plymouth.ac.uk

This is a well-established project investigating myoelectric control and assessment of teleoperated devices. Early work focussed on the control of prosthetic hands but has subsequently widened to cover generic aspects of human control of robot mechanisms, especially the modelling and assessment of assistive devices. Successful, single site, multidimensional control of a virtual robot hand has been demonstrated.

Natural Object Categorisation

Contact: Dr. Phil Culverhouse
Tel: +44 (0) 1752 233517
p.culverhouse@plymouth.ac.uk

Automatic machine identification of natural objects is important yet difficult. Man-made object recognition is simpler because design details used to built such objects may later be used to guide their identification. Natural objects lack explicit design information and are not uniform in their construction, compounding the task of determining what the object is. A programme to develop automatic psychologically-plausible methods for the identification of species of marine plankton and other natural specimens has been on-going since 1989. A prototype software tool has been under going fields trials, with EU funding, that is capable of automatically labelling marine plankton in a few seconds. Future work includes industrial application and increasing the scope of operation with overlapping specimens and highly cluttered images.

 

Robot Football

Contact: Mr. Paul Robinson
Tel: +44 (0) 1752 232595
p.robinson@plymouth.ac.uk

The design, construction, control and behaviour of intelligent co-operative agents provides opportunities for multidisciplinary research into three strategically important technologies, namely; microrobotics, real time environmental sensing (mainly vision) and artificial intelligence. International Mirosot (micro robot soccer tournament) robot football competitions provide a forum for assessing progress.

 

'Insect' digital vision

Contact: Dr. Phil Culverhouse
Tel: +44 (0) 1752 233517
p.culverhouse@plymouth.ac.uk

A new research programme is developing very low cost camera systems that 'see' the world in the same manner as insects. The software, hardware and optics are being developed to simplify real-time robot navigation and object recognition.

 

Robot Calibration

Contact: Mr. Paul Robinson
Tel: +44 (0) 1752 232595
p.robinson@plymouth.ac.uk

As software packages for industrial automation design and robot off-line programming become more widely available, interest in robot calibration has increased. Industrial robots linked to off-line programming packages require regular calibration. This project is concerned with developing practical, inexpensive methods of calibrating factory floor robots.

 

Rapid Prototyping

Contact: Mr. Paul Robinson
Tel: +44 (0) 1752 232595
p.robinson@plymouth.ac.uk

Construction of moulds for fibreglass boat hulls (know as plugs), and similar structures, is a time consuming, expensive procedure. An on-going, EU funded project, has established the feasibility of using off-line programmed robots to deliver a specially developed material in an accurately controlled format for mould construction. A demonstrator system is at the design stage.


Instruction-Based Learning for Mobile Robots (IBL)

Contact: Dr. Guido Bugmann
Tel: +44 (0) 1752 23 25 66
gbugmann@soc.plymouth.ac.uk

Future Domestic and helper robots will need to learn the special needs of their users. In this EPSRC-funded project with Edinburgh, we investigate the use of unconstrained natural language to teach new procedures to a mobile robot. For instance, how to navigate between two places in a miniature town (for more details, see the project web page: http://www.tech.plym.ac.uk/soc/staff/guidbugm/ibl/index.html.