RL-CONFORM 2023: 3rd RL-CONFORM Workshop: Reinforcement Learning meets HRI, Control, and Formal Methods Detroit, MI, United States, October 1, 2023

CALL FOR CONTRIBUTIONS

Dear colleagues,

We are thrilled to announce the 3rd RL-CONFORM: “Reinforcement Learning meets HRI, Control, and Formal Methods”, which will be held as a one-day workshop at IROS’23 on Sunday October 1, 2023. This year, we have a particular focus on contributions to make RL for robotic systems a reality, and invite you to submit your  research papers, software, lessons learned, tutorials, or field reports. Best presented contributions will have the chance to win an award supported by the RAS TC on Robot Learning. Find more details in the Call for Contributions below.

==============================

IMPORTANT DETAILS

• When: Sunday, October 1st, 2023.
• Where: Hybrid event co-located with IROS 2023 in Detroit, USA.
• Website: https://rlconform-workshop.github.io
• Contribution Submission Deadline: September 1, 2023 (AoE)
• Notification of Acceptance: September 12, 2023

Submission website: https://easychair.org/conferences/?conf=rlconform2023

• Submission format: up to 4 pages (excluding references) that describe your contribution and how its presentation will benefit the RL-CONFORM community.  Papers should be formatted in the IROS 2023 style guidelines https://ieee-iros.org/call-for-papers/
• Contact: rlconform.workshop@gmail.com 

==============================

AIM AND SCOPE

Reinforcement Learning (RL) is promising for continuous learning and discovery of optimal policies for complex tasks. However, a major open challenge is the safety and robustness of decision-making for RL systems that has involved efforts from a variety of communities, including RL, human-robot interaction (HRI), control, and formal methods (FM). 

The aim of this multidisciplinary workshop is to bring together researchers both in industry and academia from these communities to identify and clearly define key challenges and propose and debate existing approaches related to safe and robust exploration, formal safety and stability guarantees of control systems, safety in physical human-robot collaborative systems, and discuss methods and benchmarks to accelerate safe and robust RL research. To share ideas between communities, this workshop is designed to encourage fruitful and lively discussions between researchers and is open to anyone.

TOPICS OF INTEREST

Based on the target areas and the discussions during our RL-CONFORM workshop at last year’s IROS, topics of interest include but are not limited to:  

• Data-efficiency, sim-to-real gap, and guided exploration in RL; 
• Safety guarantees, shielding, invariant sets, and online verification; 
• Stability, Lyapunov functions, controllability, and model identification; 
• Query sample-efficiency, human-robot interaction, learning from demonstration, and human feedback;
• Existing and new benchmarks to accelerate safe and robust RL research.

CALL FOR CONTRIBUTIONS

Bringing RL into the real world requires various contributions from theory to software libraries. We want to represent the variety of contributions at this year’s RL-CONFORM.  We invite you to submit your contribution to enable safe, robust and efficient RL for robotic systems. Contributions could be research papers, lessons learned from implementing RL algorithms, new software, tutorials, new benchmarks, datasets, or field reports from applying RL to robot hardware. We will have three awards to recognize the presented contributions that have 1) the best poster presentation, 2) best demonstrated application, 3) best pedagogical value. Please submit up to 4 pages (excl. references) using the IROS template in which you describe your contribution to RL and explicitly motivate how the community will benefit from the presentation of your contribution at RL-CONFORM. All accepted contributions will have the opportunity to be presented at the workshop during a short teaser session and discussed in a poster session. This is a non-archival venue: there will be no formal proceedings, but we strongly encourage the authors to publish their contributions on arXiv; links to the contributions will be placed on the workshop’s website and will remain available after the workshop. Your contribution may be submitted to other venues in the future.

==============================

Invited Speakers and Panelists:

• Sheila McIlraith, University of Toronto
• Johannes Stork, Örebro University
• Matthew Gombolay, Georgia Institute of Technology
• Stefanos Nikolaidis, University of Southern California
• Daniel Brown, University of Utah
• Bei Peng, University of Liverpool
• Rika Antonova, Stanford University
• Qi Dou, The Chinese University of Hong Kong

And more to be confirmed.

====================

Organizing Committee:

• Christian Pek, Delft University of Technology
• Sanne van Waveren, Georgia Institute of Technology
• Hang Yin, University of Copenhagen