RECRUITMENT
Postdoc Position
Job Title: Postdoc position at
KAIST, South Korea
Position Summary: The position is at the Robust Intelligence and Robotics Lab (RIRO Lab, https://rirolab.kaist.ac.kr/) at the school of computing, KAIST. The lab is currently composed of 10 people including a professor, students, visiting scholars, and secretarial support. The RIRO lab aims to empower robots to be trusted collaborators in every part of our society by introducing novel learning, reasoning, and execution capabilities. We are particularly focusing on the development of collaborative mobile manipulation system with learning, planning, and communication capabilities. We have excellent facilities including a human-size mobile manipulator, small mobile manipulators, human-like dexterous hands, haptic gloves, etc. We also have strong collaborations with many domestic/international groups.
Job Description: We are looking for highly motivated postdoc, who can lead a mobile manipulation project: “Cooperative Intelligence Laboratory for Heterogeneous Robots” supported by National Research Foundation of Korea (NRF).
The project is a three-year long group research program with two perception & control research labs at KAIST. The goal of this project is to enable multiple heterogeneous robots collaboratively assemble or deliver objects.
We are looking for post-doctoral researcher who can work on some of the following:
- Development of task or task-and-motion planning algorithms for mobile manipulation
- Development of (inverse) reinforcement learning framework for in-hand manipulation
- Development of multi-agent system
- Design, implementation, and evaluation of algorithms on our experimental platforms
You will be supported by multiple graduate students with more focus on the design of algorithm and implementation. Some of our recent publications are:
[1] Shen Li*, Daehyung Park*, Yoonchang Sung*, Julie A. Shah, and Nicholas Roy, "Reactive Task and Motion Planning under Temporal Logic Specifications", IEEE Int'l. Conf. on Robotics and Automation, 2021. (ICRA2021) [PDF][Video] (*- authors contributed equally)
[2] Daehyung Park, Michael Noseworthy, Rohan Paul, Subhro Roy, and Nicholas Roy. "Inferring Task Goals and Constraints using Bayesian Nonparametric Inverse Reinforcement Learning", Conference on Robot Learning (CoRL2019)
[3] Daehyung Park*, Jacob Arkin*, Subhro Roy, Matthew R. Walter, Nicholas Roy, Thomas M. Howard, and Rohan Paul. "Multi-Modal Estimation and Communication of Latent Semantic Knowledge for Robust Execution of Robot Instructions", The International Journal of Robotics Research (IJRR), 2020.
YOUR PROFILE
In order to be successful in this position you will need to have:
- PhD in Robotics, AI, or closely related field
- Good academic track record with publications
- Strong background in one of the following: Planning, Reinforcement Learning, Computer Vision, Deep Learning, Control
- Good English skills, written and spoken
- Good skills in Python, Deep Learning / Reinforcement Learning Programming (e.g., PyTorch and/or Tensorflow, etc.) and Linux
- Experties in working with real robot systems is a big plus
The following qualifications would be an advantage:
- Familiarity with state-of-the-art task planning and/or task-and-motion planning
- Experience using PDDL or Temporal Logic
- Ability to work with people in your research and try to being involved in research of others to multiply the effect of collaborations.
- Have a clear open and transparent communication attitude.
- Be highly motivated to achieve excellence in the interest of science and engineering innovation within the university, the institutes, the RIRO lab and for yourself.
Application Process and Agenda
The position is to be filled as soon as possible, with an ideal starting date of March 1, 2022. Applications will be considered on a rolling basis. The initial term is one year, with a possible extension depending on performance.
More information? See here:
Recruiting Officer Information
| Person in Charge | Dr. Daehyung Park | Contact Information | Contact Information is confidential | E-mail is confidential |
|---|