To get updates from our aUToronto self-driving car team, please visit this page.
Dynamic Systems Lab News
Have a new idea for safe learning and decision making? We are organizing a NeurIPS'21 Workshop on Deployable Decision Making in Embodied Systems (DDM) and are inviting short paper/abstract submissions (deadline Oct. 15). Check our webpage for details www.dynsyslab.org/deployable-decision-making-in-embodied-systems/
Our IROS workshop on Safe Robot Autonomy is only a few days away! Join us live next Monday (Sept. 27) 10:10 am ET at the Youtube Livestream tiny.cc/safe-robotics-panel. We hope to see you all and await your questions for our panelists here tiny.cc/safe-robotics-panel-q!
Mark you calendars and start preparing! Our IROS’21 Workshop on Safe Real-World Robot Autonomy is coming to you on Mon Sep 27, 10am to noon ET. For now, check out the invited spotlight talks on recent advances in safe decision making for robotics.
IROS 2021 Workshop on Safe Real-World Robot Autonomy
www.dynsyslab.org
Interdisciplinary Discussion on Requirements, Approaches, and Benchmarks for Safe Real-World Robot AutonomyJoin us on September 27 at the IROS Workshop on Safe Real-World Robot Autonomy! Check out our program and add your questions for the panelists here: www.dynsyslab.org/safe-real-world-robot-autonomy/ Early registration closes September 5!
Our review paper on **Safe Learning in Robotics** is out, including open-source code. Wondering how model-driven and data-driven approaches can work together to achieve safe learning? Check out the paper! We review approaches from both the controls and the reinforcement learning communities. arxiv.org/abs/2108.06266
We will post more snippets of the paper over the coming weeks 🙂
--
University of Toronto Engineering
We are organizing a NeurIPS 2021 workshop on Deployable Decision Making in Embodied Systems (DDM). Check out the exciting lineup of speakers and how you can contribute workshop papers: www.dynsyslab.org/deployable-decision-making-in-embodied-systems/
Robotic systems are expected to guarantee high performance despite being deployed in unknown and dynamic environments. We combine robust model predictive control with L1 adaptive control for guaranteed fast and accurate stabilization in the presence of model uncertainties.
Paper: www.dynsyslab.org/wp-content/papercite-data/pdf/pereida-ijrnc21.pdf
Fusion of data from ultrawideband (UWB) and inertial sensors can provide accurate positioning in GPS-denied areas. Reliable fusion requires calibration of the spatio-temporal offsets. Our work on online calibration of UWB-aided inertial localization systems was accepted at IROS 2021. Paper: www.dynsyslab.org/wp-content/papercite-data/pdf/goudar-iros21.pdf
If you are at the CIFAR DLRL summer school, see Prof. Angela Schoellig's talk today about safe robot learning. She will also introduce our Safe Robot Learning Survey that just got accepted. Subscribe here for early access to the survey: tiny.cc/safe-learning.
dlrl.ca
Where the world comes to learn AI Learn more Every year, the CIFAR Deep Learning + Reinforcement Learning (DLRL) Summer School brings together graduate students, post-docs and professionals to cover t...There is an exciting workshop at ICML today on Reinforcement Learning (RL) for Real Life. Prof. Angela Schoellig will be part of a panel on RL and Robotics with other amazing fellow roboticists. Join us at 1-2PM EDT, July 23.
