To get updates from our aUToronto self-driving car team, please visit this page.
Dynamic Systems Lab News
Monday November 30th, 2020
ICYMI: Check out our IROS2020 Presentation about playing catch with a mobile manipulator. We proposed a planning and control framework that achieves a success rate of 85%.
Paper: arxiv.org/pdf/2003.07489.pdf
Catch the Ball: Accurate High-Speed Motions for Mobile Manipulators via Inverse Dynamics Learning
In this project, we focus on the mobile manipulator ball-catching task. We introduce a framework including two components: a bi-level motion optimization alg...Wednesday November 25th, 2020
ICYMI: The IROS2020 presentation for our RA-L paper. We learn robot parameters for trajectory estimation in a Gaussian variational inference setting. Our method works without ground truth, i.e. with only noisy measurements including outliers
Paper: www.dynsyslab.org/wp-content/papercite-data/pdf/wong-ral20b.pdf
Wednesday October 7th, 2020
How is robotics being used to in the era of COVID-19? The University of Toronto Robotics Institute's faculty have produced a white paper regarding the role of Robotics in the pandemic response and beyond.
arxiv.org/abs/2009.08577
Friday October 2nd, 2020
Check out the University of Toronto Engineering News article about aUToronto's win: uofteng.ca/sdOCUg
Thursday October 1st, 2020
By exploiting differential flatness and the properties Gaussian Processes, we develop a learning-based controller that achieves high-accuracy tracking while guaranteeing safety. Find our upcoming LCSS paper: ieeexplore.ieee.org/document/9140024
Tuesday September 22nd, 2020
With the new semester having started, we want to welcome our new graduate students: Lukas, Shichen, Justin, Xintong, and Jingxing! See the whole team: www.dynsyslab.org/team/
Tuesday September 8th, 2020
Want to play catch with a mobile manipulator? Why not! We proposed a planning and control framework that achieves a success rate of 85%, accepted to IROS2020.
Paper: www.dynsyslab.org/wp-content/papercite-data/pdf/dong-iros20.pdf
Friday August 14th, 2020
Prof. Angela Schoellig presented "Safe and Efficient Active Learning Strategies for Robotics Applications" at the ICML workshop on Real World Experiment Design and Active Learning check out the video at 7:32:00!
icml.cc/virtual/2020/workshop/5745
Wednesday August 12th, 2020
Wondering how to guarantee success in cross-robot experience transfer? Our L-CSS paper presents an approach to predict transfer performance in cross-robot model transfer and highlights the asymmetric nature of cross-robot transfer problems.
Paper: arxiv.org/pdf/2006.16126.pdf
