Welcome to the news blog of the EPSRC Centre for Doctoral Training in Future Autonomous Robotic Systems (FARSCOPE). Please also visit our main web page at http://farscope.bris.ac.uk and follow us on Twitter. FARSCOPE is based at the Bristol Robotics Laboratory, a partnership between the University of Bristol and the University of the West of England.

Thursday, 29 March 2018

Three Extra Places in FARSCOPE (and more coming...)

Three great ways are now available to join the FARSCOPE Centre for Doctoral Training in September 2018, all mixing the broad experience of FARSCOPE with the opportunity to work with an industry partner on real-world robotics problems.  All three share a common theme of multi-agent systems for industry use, but in different places and scales.

Develop next generation swarm robots with Toshiba

Join the Swarm Robotics team with supervisor Sabine Hauert and help develop swarm robots, fantastic for their robustness and flexibility but incredibly hard to engineer, for industrial applications.  How would ants run a factory or a warehouse...?  More information and details of how to apply are here.

Help steer future shipping with Thales

Current ships obey the COLREG rules to avoid collisions, and future autonomous ships will need to be behave similarly when they're introduced.  But the COLREGs only cover pairs of vessels - that makes them easy to understand, but they can be limiting or even dangerous in crowded situations.  So how can we write three-way rules?  How would a computer understand them?  And how would we humans understand them?  What would the best rules look like, and for how many ships at a time?  When all AI has to be able to explain itself, how do you trade off performance for ease of understanding?  Join supervisors Jonathan Lawry and Arthur Richards to model and help shape future rules-of-the-sea.

Engineer 'Hybrid Autonomous Systems' with T-B PHASE

T-B PHASE is the Thales-Bristol Partnership on Hybrid Autonomous Systems Engineering, an EPSRC / Thales Partnership for Prosperity.  'Hybrid' means these autonomous systems don't get a cushy life in the lab - they have to interact with the real world, its uncertainty, and the people in it.  This is a really broad PhD opportunity and you'll get to help shape it over your first year.  The spec is to do something around these challenges:

  • Distributed decision-making: how do we engineer, deploy and support systems without a central decision-making authority?  How can the interaction of those systems be managed?  How do we consider emergent behaviour of the system as a whole?
  • Cascade failure: how do we predict how failures affect other parts of large systems?  How do we prevent knock-on effects in which local failures lead to large-scale problems?  How can we confidently engineer systems that are strongly interconnected?
  • Life course autonomy: how do we design systems to remain in service for decades?  How do we account for changing demands, upgrades, and evolving user behaviour?  How do we engineer something so our great-grandchildren can still make use of our work?
  • Human factors: how can autonomous systems be operated by and with humans?  How do we combine the untiring efficiency of machines with the creative resilience of humans?  How will humans cope with a machine that surprises them with its capabilities?

And you'll need to relate it back to one or more of these applications:

  • Traffic management for small unmanned air vehicles (“drones”)
  • Future railway traffic management and signalling
  • Semi-autonomous and robotic search and rescue

T-B PHASE has a range of supervisors to work with, depending on how your project develops.  These are some of the hardest questions in the business, making the difference between a cool lab demos and capabilities that can actually earn their keep in the wild.

How to get involved

You can apply for any or all of these together.  The deadline is 4th May, but we may have further places coming on-stream after that.  Submit your application form through the online system via http://www.bristol.ac.uk/study/postgraduate/apply/ and state in the "Funding" section which projects you want to be considered for.  For the "next generation swarm robots" project, please email Sabine Hauert first (see her page for details).