Role type: Full-time, fixed term for 12 months (possibility of extension)
Salary: Level A – $77,171 – $104,717 p.a. plus 17% super, minimum salary for PhD graduate - $97,558
The Research Fellow will work on reinforcement learning, or other data-driven control methodologies, with a view to ensure constraint satisfaction during both the learning and deployment phases to avoid unsafe behaviour.
The use of learned models and controllers is becoming prevalent. Online learning can create adaptable algorithms for control in unknown or time-varying environments or for plug-and-play interaction with other systems. However, criticality of applications, such as autonomous driving and flight, demands constraint satisfaction during the training to still guarantee safe operation while adapting the controllers online. In this position, you will explore, develop, analyse, and demonstrate various approaches in safe reinforcement learning for linear and nonlinear dynamic models. Extensions to multi-agent planning and control will be highly desirable.
There is funding available for attending international conferences and visiting an internationally recognized research group for collaboration on the topic of this position. The research group has longstanding collaboration links with researchers from the University of California Irvine, University of Illinois Urbana-Champaign, Princeton University, Australian National University, and KTH Royal Institute of Technology.
Responsibilities include:
- Providing expertise in design, analysis, and demonstration of reinforcement learning and data-driven control with constraint satisfaction for safety.
- Publication of high-quality research papers on the topic of reinforcement learning and data-driven control in peer-reviewed top journals and conferences in computer science and control engineering.
- Providing project representation to various internal or external stakeholders.
- Attending and participating in relevant meetings, seminars, and conferences.
- Contributing to training, mentoring, and supervision of students.
About You
You are a confident communicator with well-developed interpersonal and negotiation skills with the ability to build and maintain relationships with internal and external stakeholders within a diverse work environment. You are organised, detail oriented with a strong work ethic, commitment to continuous improvement, openness to new ideas and creative approaches to problem solving within established timelines. Ideally, you will further have:
- A PhD in Engineering, Computer Science, or Applied Mathematics.
- A solid understanding of the theory of at least of one of the topics of reinforcement learning, optimal control, optimization theory, or system identification.
- An excellent track record of quality research as evidenced by publications in the top peer-reviewed journals and conferences of systems and control, signal processing, computer science, or applied mathematics.
Please take look at the job advertisement and position description in the link below:
https://jobs.unimelb.edu.au/caw/en/job/909399