Probabilistic Robotics In the age of Deep Learning

IROS 2023 Workshop | Date: TBD | Room: TBD

2022 / 2023


Fabio Ramos
Michael Kaess
Ayoung Kim
Andreas Kruse
Fabio Ramos Michael Kaess Ayoung Kim Andreas Krause
NVIDIA & University of Sydney (USA and Australia) Carnegie Mellon University (USA) Seoul National University (Korea) ETH Zürich (Switzerland)
Talk: TBD
Abstract: ↵


Talk: TBD Talk: Deep learning with a graph SLAM for robot manipulation Talk: Safe and Efficient Exploration in Bayesian Model-based Reinforcement Learning
Sharon Yixuan Li
Hermann Blum
Ali Agha
Sharon Yixuan Li Hermann Blum Ali Agha
University of Wisconsin (USA) ETH Zürich (Switzerland) NASA JPL & Caltech (USA)
Talk: How to Handle Data Distributional Shifts? Challenges, Opportunities and Path Forward Talk: 4 years of fishyscapes: uncertainty estimation in the wild Talk: Uncertainty-aware Robotic Autonomy in Extreme Environments


Time Type Speaker
08:30 - 09:15 Talk Michael Kaess
09:15 - 10:00 Talk Ayoung Kim
10:00 – 10:15 Coffee Break  
10:15 - 11:00 Talk Fabio Ramos
11:00 - 11:45 Talk Hermann Blum
11:45 - 13:00 Lunch  
13:00 – 13:45 Talk Andreas Kruse
13:45 – 14:30 Talk Sharon Yixuan Li
14:30 – 14:45 Coffee Break  
14:45 - 15:30 Talk Ali Agha
15:30 - 16:15 Paper Presentations & Posters Norman Marlier
16:15 – 16:30 Coffee Break  
16:30 - 17:30 Panel Discussion All the speakers

Accepted Papers


Robots are envisioned to perform complex tasks in real, noisy, imperfect, partially observable, hardly perceivable and unexpectedly behaving environments. Due to the nature of the physical world, it is crucial to endow robotic systems with the ability to handle uncertainties. Probabilistic robotics, a vibrant field that has gained enormous popularity since its inception in the mid-1990s, provides a compelling paradigm to address this challenge.

The evolution of probabilistic techniques in robotics connects the field to other adjacent research domains such as artificial intelligence (AI). While classical AI still plays an important role in robotic systems, its modern counterpart - empowered by Deep Learning - is able to remarkably enhance the functionality of our robots with its impressive performance on different robotic tasks. However, each coin has two sides. The uninterpretable nature of DL and its data-hungry learning process unfortunately limits the applications of various probabilistic approaches. Though there have been advances in pursuing extension of this concept in recent years, many core challenges associated with robot deployment to the real world still remain unsolved. In consideration of this, it is important to take a look at the probabilistic techniques in the age of pre-DL and peri-DL from a probabilistic robotics point of view.

In light of the above, this workshop aims at providing a forum for robotic and machine learning researchers as well as industrial experts with experience on developing probabilistic methods. The discussions will be centered on past achievements, current requirements, urgent challenges and future research directions and promising applications.


Janis Postels
Jianxiang Feng
Jongseok Lee
Matthias Humt
Yizhe Wu
Rudolph Triebel
Janis Postels Jianxiang Feng Jongseok Lee Matthias Humt Yizhe Wu Rudolph Triebel
ETH Zurich (Switzerland) DLR (Germany) DLR (Germany) DLR (Germany) University of Oxford (United Kingdom) DLR and TU Munich (Germany)



Should you have any questions, please do not hesitate to contact the organizing committee.