Postgraduate research opportunities

Robotics – Multi-Agent Planning

The research scope of “Multi-Agent Planning” has at its centre the study and development of multi-agent task and path planner that also identifies the appropriate type of perception, mapping and localisation methods for varying tasks, application context and dynamic environment condition.

Number of places

1

Funding

Home fee, Stipend

Opens

11 February 2020

Deadline

30 April 2020

Duration

4 years

Eligibility

Qualifications and Experience

  • Master degree in a relevant field (i.e. AI-Machine Learning, robotics, software engineering)
  • Strong theoretical and practical background in AI and/or optimisation techniques, planning domain description, ontology based knowledge representation, semantic and probabilistic reasoning applied to multi-robot mission planning paradigms
  • Strong programming skills (C++, Python) on Linux environment
  • Experience with SW development in ROS and simulation in Gazebo
  • Understanding of perception, mapping and localisation solutions for robot navigation

Experience with or knowledge of a selection of the following will be considered as an asset:

  • Machine Learning techniques (i.e. Bayesian networks, Deep Learning)
  • Analytical or AI-based image processing, state estimation or filtering
  • OWL, PDDL, MAPL, TensorFlow, OpenCV, PCL, Qt and Web technologies
  • GitLab and Continuous Integration methodologies
  • Proficiency with technical documentation.

The successful candidate will have the following skills

  • A proactive approach, with initiative and ability to work independently
  • Ability to:
  • Synthesise, summarise and draw conclusions
  • Adhere to strict standards of confidentiality
  • Work in distributed international teams
  • Strength to cope with schedules and deadlines
  • Excellent organisational and communication skills
  • Excellent written and spoken English.

Applications from Home, Rest of UK and EU students will receive funding of the home fee and stipend.

Find out more about this exciting PhD opportunity by clicking through the tabs above.

Individuals interested in this project should email dmem-pgr@strath.ac.uk, along with the title of project you are applying for and attach your most up-to-date cv.

Project Details

Are you looking to gain direct industrial experience whilst developing your specialist research area of interest?

Space Applications Services and the University of Strathclyde (UK) are offering a joint academic-industrial PhD.  The successful doctorate will be based for 1.5 years at the University of Strathclyde and 2.5years in our main office at Sint-Stevens-Woluwe (Brussels Area).  The candidate will mainly report to the PhD supervisors from both organisations.

Space Applications Services is based in the Brussels area that provides products and services for the space sector in many areas ranging from Avionics, Robotics, Human Exploration, Science and Earth Observation.

The research scope of “Multi-Agent Planning” has at its centre the study and development of multi-agent task and path planner that also identifies the appropriate type of perception, mapping and localisation methods for varying tasks, application context and dynamic environment condition. First, the planner will be applied as a part of an offline planning engine within a multi-robot control centre. The outcome will be tested in an online planning context and deployed on robots, empowering heterogeneous teams of robots to organise themselves logically for cooperative perception and task management.

Tasks and Responsibilities

  • Reviewing the state-of-the-art in multi-agent planning based on AI or optimisation approaches solutions and commonly used perception, mapping and localisation techniques for navigation of underwater and space robotics; identifying the exact scope of the PhD research tasks
  • Investigating/adapting simulation environments for prototyping and testing novel concepts related to planned coupled with sensor fusion
  • Developing and testing mission planning tasks and path planning for multi-agent systems, closely integrated with navigation methods
  • Studying and fine-tuning data fusion algorithms allowing the robots to map and navigate autonomously the unknown environment, e.g. agriculture or mining fields
  • Integrating, testing and optimising the planning method with a command and control centre and on a target robot architecture
  • Generating the technical documentation required all along the process.

Funding Details

This fully-funded industrial PhD opportunity will cover Home and EU Fees and Stipend.

We will accept applications from international students who can confirm in their email application that they are able to pay the difference between the Home and International fees (approximately £15,009 per annum). If you are unable to cover this cost the application will be rejected.

Supervisor

Space Applications Services and the University of Strathclyde (UK) are offering a joint academic-industrial PhD.  The successful doctorate will be based for 1.5 years at the University of Strathclyde and 2.5years in our main office at Sint-Stevens-Woluwe (Brussels Area).  The candidate will mainly report to the PhD supervisors from both organisations.

The University of Strathclyde supervisor for this PhD is Professor Xiu Yan.   Please note: We request that potential candidates do not contact Professor Yan and instead direct all questions to dmem-pgr-recruitment@strath.ac.uk to the attention of Dr Dorothy Evans.

Further information

Space Applications Services and the University of Strathclyde (UK) are offering a joint academic-industrial PhD.  The successful doctorate will be based for 1.5 years at the University of Strathclyde and 2.5years in our main office at Sint-Stevens-Woluwe (Brussels Area).  The candidate will mainly report to the PhD supervisors from both organisations.

Space Applications Services is based in the Brussels area that provides products and services for the space sector in many areas ranging from Avionics, Robotics, Human Exploration, Science and Earth Observation.

The research scope of “Multi-Agent Planning” has at its centre the study and development of multi-agent task and path planner that also identifies the appropriate type of perception, mapping and localisation methods for varying tasks, application context and dynamic environment condition. First, the planner will be applied as a part of an offline planning engine within a multi-robot control centre. The outcome will be tested in an online planning context and deployed on robots, empowering heterogeneous teams of robots to organise themselves logically for cooperative perception and task management.

PhD Tasks and Responsibilities

  • Reviewing the state-of-the-art in multi-agent planning based on AI or optimisation approaches solutions and commonly used perception, mapping and localisation techniques for navigation of underwater and space robotics; identifying the exact scope of the PhD research tasks
  • Investigating/adapting simulation environments for prototyping and testing novel concepts related to planned coupled with sensor fusion
  • Developing and testing mission planning tasks and path planning for multi-agent systems, closely integrated with navigation methods
  • Studying and fine-tuning data fusion algorithms allowing the robots to map and navigate autonomously the unknown environment, e.g. agriculture or mining fields
  • Integrating, testing and optimising the planning method with a command and control centre and on a target robot architecture
  • Generating the technical documentation required all along the process.

Applicant Qualifications and Experience

  • Master degree in a relevant field (i.e. AI-Machine Learning, robotics, software engineering)
  • Strong theoretical and practical background in AI and/or optimisation techniques, planning domain description, ontology based knowledge representation, semantic and probabilistic reasoning applied to multi-robot mission planning paradigms
  • Strong programming skills (C++, Python) on Linux environment
  • Experience with SW development in ROS and simulation in Gazebo
  • Understanding of perception, mapping and localisation solutions for robot navigation

Experience with or knowledge of a selection of the following will be considered as an asset:

  • Machine Learning techniques (i.e. Bayesian networks, Deep Learning)
  • Analytical or AI-based image processing, state estimation or filtering
  • OWL, PDDL, MAPL, TensorFlow, OpenCV, PCL, Qt and Web technologies
  • GitLab and Continuous Integration methodologies
  • Proficiency with technical documentation.

The successful candidate will have the following skills

  • A proactive approach, with initiative and ability to work independently
  • Ability to:
  • Synthesise, summarise and draw conclusions
  • Adhere to strict standards of confidentiality
  • Work in distributed international teams
  • Strength to cope with schedules and deadlines
  • Excellent organisational and communication skills
  • Excellent written and spoken English.

What Do We Offer?

  • Conducting research as part of a vibrant robotics and autonomous research team within Space Mechatonics Systems Technology Lab (SMeSTech) at the University of Strathclyde and working in a growing company with Staff located in Belgium, Germany and the Netherlands
  • A professional, pleasant atmosphere with motivated and passionate staff, where autonomy and initiatives are encouraged
  • Four year PhD scholarship and support from all involved parties in conducting state of the art research in robotics

Contact us

How to apply

Please send your CV and Motivation Letter (in English) to dmem-pgr-recruitment@strath.ac.uk to the attention of Dr Dorothy Evans.

The candidate shall be from the EU or UK member states.

We are looking for someone to start as soon as possible.