- Visiting_residents - Australian Centre for Robotics

[Everybody at ACFR] ACFR Seminar Series: Zeyu Zhang (UC Berkeley) - Towards General Embodied Reasoning
by Sabrina Asri 23 Jul '26

23 Jul '26

Faculty of Engineering ACFR Seminar Series [https://d31hzlhk6di2h5.cloudfront.net/20260723/44/a0/ac/85/83de99ae1318da0a…] [https://d31hzlhk6di2h5.cloudfront.net/20260723/90/62/a1/a0/e806283be685b4d0…] Towards General Embodied Reasoning Presented by Zeyu Zhang (UC Berkeley) Date: Friday, 24 July 2026, at 2:00 pm AEST Venue: ACFR seminar area, J04 level 2 (Rose St Building<https://url.au.m.mimecastprotect.com/s/HNKJCBNqjlCMGgDLxfMcNf2WyeD?domain=t…>) Zoom ID: https://uni-sydney.zoom.us/s/87306457455<https://url.au.m.mimecastprotect.com/s/Sw1TCD1vlpTn9Ro87CxfRfj35gq?domain=t…> Abstract: Building general-purpose embodied agents requires moving beyond specialized systems toward models that can understand, reason about, and interact with diverse physical environments. In this talk, Zeyu Zhang will present his recent research on general embodied reasoning, covering 3D scene understanding, long-horizon navigation, robotic control, domain adaptation, and self-evolving agents. The talk will discuss how multimodal pretraining, scalable synthetic data, and data-centric learning can produce transferable representations that support a wide range of tasks, including captioning, grounding, question answering, dialogue, spatial reasoning, and planning. It will then examine how environmental understanding can be extended to embodied decision-making through hierarchical frameworks that combine fast reactive control with slower long-horizon reasoning. Zeyu will also discuss how carefully designed post-training, parameter-efficient adaptation, and verifiable feedback can improve navigation and control while preserving previously acquired perception and reasoning capabilities. Finally, the talk will explore how embodied agents can continue improving when manually specified objectives are incomplete or inaccurate, using self-supervised feedback, world-model-based exploration, adaptive memory, and stage-aware learning signals. Together, these studies highlight the importance of scalable data, spatially grounded foundation models, restrained reinforcement learning, and continual self-improvement for advancing toward general embodied intelligence. Bio: Zeyu is an incoming PhD student at UC Berkeley BAIR, advised by Professor Pieter Abbeel, Professor Alexei Efros, and Professor Angjoo Kanazawa. He received his bachelor’s degree from the Australian National University, where he was advised by Professor Richard Hartley and Professor Ian Reid. [Register for Event >]<https://url.au.m.mimecastprotect.com/s/_sewCE8wmrtg7QlZwfDhyf7aYCr?domain=t…> [https://d31hzlhk6di2h5.cloudfront.net/20260723/2e/8d/c7/1c/1c70baef8b350e38…] [Web Site]<https://url.au.m.mimecastprotect.com/s/eU4VCJyBrGf1J7BjkS7s9fypLgI?domain=t…> [LinkedIn]<https://url.au.m.mimecastprotect.com/s/_JIHCK1DvKTBojDRQH8twf5iM3k?domain=t…> [X]<https://url.au.m.mimecastprotect.com/s/alOaCL7EwMfmWJkz0CXuMfy6EFv?domain=t…> [YouTube]<https://url.au.m.mimecastprotect.com/s/XaPICMwGxOtRrQ2nphoCKf812z9?domain=t…> Copyright © 2026 The University of Sydney Please add acfr.admin(a)sydney.edu.au<mailto:acfr.admin@sydney.edu.au> to your address book or senders safe list to make sure you continue to see our emails in the future. Cricos Code 00026A TEQSA PRV12057 Disclaimer<https://url.au.m.mimecastprotect.com/s/_edYCNLJyQUEW2ZgYtwFDfyfNP1?domain=t…> | Privacy statement<https://url.au.m.mimecastprotect.com/s/A02yCOMKzVT086NBDt8HzfGa1So?domain=t…> | _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] Lab visit 2.30pm
by Thien Hoang Nguyen 23 Jul '26

23 Jul '26

Hi everyone, Apologies for the late notice. I will take a visitor from VinUni to J19 at around 2.30pm - 3.30pm. We will stay mostly at the Ag area after a quick tour around the lab. Please hide your stuffs at appropriate. Thank you. Best regards, Thien Nguyen | Vonwiller Research Associate | Australian Centre for Robotics Sydney Robotics Hub (J18), 8 Little Queen St, Chippendale NSW 2008 _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] Today: ACFR Seminar Series: Asst. Professor Jake Welde (Cornell) & Asst. Professor Thai Mai Thanh (VinUniversity)
by Australian Centre For Robotics Administration 23 Jul '26

23 Jul '26

Hi everyone, A kind reminder to today's seminar that will take place in J04. A light lunch will be served from 12.30pm. Further information is as below: Faculty of Engineering ACFR Seminar Series [https://d31hzlhk6di2h5.cloudfront.net/20260720/44/a0/ac/85/83de99ae1318da0a…] [https://d31hzlhk6di2h5.cloudfront.net/20260720/90/62/a1/a0/e806283be685b4d0…] 1) Symmetry and Beyond: Towards Efficient, Generalizable Learning for Low-Level Control of Aerial and Space Robots. 2) Advancing Healthcare Through Soft Robotics and Intelligent Systems Presented by Asst. Professor Jake Welde (Cornell) & Asst. Professor Thai Mai Thanh (VinUniversity) Date: Thursday, 23 July 2026, at 1:00 pm AEST Venue: ACFR seminar area, J04 level 2 (Rose St Building<https://url.au.m.mimecastprotect.com/s/W2SRCzvkyVCxZ2xnqHgizf9MaBD?domain=t…>) Zoom ID: https://uni-sydney.zoom.us/s/87306457455<https://url.au.m.mimecastprotect.com/s/dPCVCANpgjCnXWnVOFYsGfGiiJI?domain=t…> **RSVP by COB Tuesday 21 July for catering purposes. Accept this calendar invitation if you plan to attend the seminar in person and join us for a light lunch at approximately 12.30pm. Abstract: Symmetry and Beyond: Towards Efficient, Generalizable Learning for Low-Level Control of Aerial and Space Robots To perform useful, interactive work, aerial and space robots must evolve beyond simple, single-body designs (e.g., quadrotors) towards high-dimensional, articulated morphologies. However, explicit analytical controllers for such systems are not easily crafted, and numerical optimal control algorithms quickly overwhelm the lightweight processors available onboard. Reinforcement learning frontloads these burdens to offline training, typically employing vast computational resources to run brittle training pipelines fine-tuned to particular tasks. To mitigate these downsides, we exploit the natural symmetries of robotic systems to efficiently learn high-performance, generalizable policies for low-level robot control. In particular, we develop a theory of symmetry reduction for tracking control problems, proving that a tracking controller trained in a reduced setting will perform equally well on the original system while also generalizing automatically to unseen trajectories. However, the classical notion of symmetry is too rigid to apply to many practical systems, which may enjoy only a very small symmetry group. We thus propose a relaxed notion of symmetry (termed "weak invariance") that balances structure with generality, enabling us to factor out a much larger symmetry group while still attaining the same performance guarantees, greatly expanding the applicability and impact of these methods. Advancing Healthcare Through Soft Robotics and Intelligent Systems This talk presents recent advances in soft robotics and intelligent systems aimed at transforming healthcare and improving the quality of life. It highlights the development of flexible, bio-inspired robotic technologies for minimally invasive surgery, including soft robotic endoscopic platforms integrated with in situ 3D bioprinting and real-time navigation. The talk also explores wearable soft sensors and robotic devices for rehabilitation, enabling low-cost, accessible, and data-driven patient recovery. By combining soft materials, embedded systems, and artificial intelligence, these innovations offer safer, more adaptive, and patient-centered solutions. The presentation concludes with a vision for deploying such technologies in resource-constrained settings to address global healthcare challenges and enhance quality of life at scale. Bio: Jake Welde Jake Welde is an Assistant Professor with the Sibley School of Mechanical and Aerospace Engineering at Cornell University, where he leads the Geometry, Design, and Control Laboratory (or the "GeoDesiC Lab"). Prior to joining Cornell, he spent a decade at the University of Pennsylvania, where he completed his undergraduate and doctoral degrees in Mechanical Engineering and Applied Mechanics and his Masters in Robotics, working in the GRASP Laboratory with Vijay Kumar. He is broadly interested in how mathematical structure - symmetry, hierarchy, and mechanics - can be harnessed to create more capable algorithms for robot control, and how the interplay between control and morphology mediates overall robot capabilities. Thai Mai Thanh Dr. Thai Mai Thanh is an Assistant Professor of Mechanical Engineering at VinUniversity and Principal Investigator (PI) of the VinUni Biorobotics Lab. He earned his B.E. from Ho Chi Minh City University of Technology (2016), M.S. from KAIST (2019), and Ph.D. in Biomedical Engineering from UNSW Sydney (2023), where he also completed a postdoctoral fellowship in medical robotics. His research focuses on advancing healthcare through soft robotics, intelligent systems, and medical devices, with applications in minimally invasive surgery, in situ bioprinting, rehabilitation, and wearable haptics. He has authored over 40 publications and holds multiple international patents, with a strong emphasis on translating cutting-edge technologies into real-world clinical impact. [Register for Event >]<https://url.au.m.mimecastprotect.com/s/OrhLCBNqjlCMr4MAKTWtNf2V8fl?domain=t…> [https://d31hzlhk6di2h5.cloudfront.net/20260720/2e/8d/c7/1c/1c70baef8b350e38…] [Web Site]<https://url.au.m.mimecastprotect.com/s/CkOKCE8wmrtg2DgnxFBCyf7yy0x?domain=t…> [LinkedIn]<https://url.au.m.mimecastprotect.com/s/lNXUCGv0oyC0RM0q4UYFPfB3x4Y?domain=t…> [X]<https://url.au.m.mimecastprotect.com/s/WXWyCJyBrGf1vY1ywhkH9fy1cjM?domain=t…> [YouTube]<https://url.au.m.mimecastprotect.com/s/wIlDCK1DvKTBNnB9WuQIwf54w3W?domain=t…> Copyright (c) 2026 The University of Sydney Please add acfr.admin(a)sydney.edu.au<mailto:acfr.admin@sydney.edu.au> to your address book or senders safe list to make sure you continue to see our emails in the future. Cricos Code 00026A TEQSA PRV12057 Disclaimer<https://url.au.m.mimecastprotect.com/s/M2IlCL7EwMfmr1mX9TASMfyBQf_?domain=t…> | Privacy statement<https://url.au.m.mimecastprotect.com/s/9ghECMwGxOtR1ZR93s9TKf8DTfR?domain=t…> | Best regards, Sabrina Sabrina Asri | Senior Admin Officer The University of Sydney Faculty of Engineering, Australian Centre for Robotics Rose Street Building J04 | The University of Sydney | NSW | 2006 +61 2 9351 7126 | sabrina.asri(a)sydney.edu.au<mailto:sabrina.asri@sydney.edu.au> | sydney.edu.au<http://sydney.edu.au/> CRICOS 00026A TEQSA PRV12057 This email plus any attachments to it are confidential. Any unauthorised use is strictly prohibited. If you receive this email in error, please delete it and any attachments. Please think of our environment and only print this email if necessary. _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] Invitation to audit: foundations of imaging
by Donald Dansereau 22 Jul '26

22 Jul '26

The last few years I've run a short lecture series on the foundations of imaging during the winter break. This year I have the pleasure of presenting the material as a final-year elective for mechatronics students. I'd like to invite anyone with an interest in auditing this unit to get in touch. Lectures introduce computational imaging with minimal prerequisites. They cover: * Physics of image formation: light, lenses, and pixels * Generalised imaging systems: cameras, LiDAR, radar, and sonar * Multi-dimensional signals, multiplexing, frequency-domain analysis, and filtering * Coded aperture, flutter shutter, focal sweep, and light-field cameras * Machine learning and optimisation-based reconstruction methods * Modern photorealistic representations: radiance field approximation via NeRFs and Gaussian splatting By the end of the unit, you will be able to: * Explain computational imaging as the integration of optics, sensors, and computation * Use a conventional camera to measure your pulse and image around corners * Explain multi-lens systems and why we sometimes don't need lenses at all * Create your own relightable models using a monitor and a webcam * Build lensless, single-pixel, and other unconventional imaging systems Lectures can be attended in person and will be available as recordings. As an auditor you will not be attending the lab sessions but will have access to the online resources. Lectures start Aug 4th. If you're interested, reach out by return email. Sincerely, Don Donald Dansereau | Senior Lecturer The University of Sydney | Australian Centre for Robotics School of Aerospace, Mechanical and Mechatronic Engineering Rose St. Bldg. J04 | The University of Sydney | NSW | 2006 donald.dansereau(a)sydney.edu.au<mailto:donald.dansereau@sydney.edu.au> | roboticimaging.org<http://roboticimaging.org/> _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] Fw: ACFR Seminar Series: Asst. Professor Bowen Yi - Input–Output Data-Driven Control of Continuous-Time Systems
by Ian Manchester 22 Jul '26

22 Jul '26

Starting now! Get Outlook for Android<https://aka.ms/AAb9ysg> ________________________________ From: Australian Centre For Robotics Administration Sent: Tuesday, 21 July 2026 15:15:45 To: Australian Centre For Robotics Administration <acfr.admin(a)sydney.edu.au>; everybody <everybody(a)acfr.usyd.edu.au>; students(a)acfr.usyd.edu.au <students(a)acfr.usyd.edu.au>; seminars(a)acfr.usyd.edu.au <seminars(a)acfr.usyd.edu.au>; 'academics-meeting(a)acfr.usyd.edu.au' <academics-meeting(a)acfr.usyd.edu.au> Cc: Ian Manchester <ian.manchester(a)sydney.edu.au> Subject: ACFR Seminar Series: Asst. Professor Bowen Yi - Input–Output Data-Driven Control of Continuous-Time Systems When: Wednesday, July 22, 2026 1:00 PM-2:00 PM. Where: ACFR J04 Level 2 Seminar Area:; https://maps.app.goo.gl/PUJdbB9oUTtg5fLr8; ZOOM:; https://uni-sydney.zoom.us/s/87306457455 Faculty of Engineering ACFR Seminar Series [https://d31hzlhk6di2h5.cloudfront.net/20260721/44/a0/ac/85/83de99ae1318da0a…] [https://d31hzlhk6di2h5.cloudfront.net/20260721/90/62/a1/a0/e806283be685b4d0…] Input–Output Data-Driven Control of Continuous-Time Systems Presented by Asst. Professor Bowen Yi Date: Wednesday, 22 July 2026, at 1:00 pm AEST Venue: ACFR seminar area, J04 level 2 (Rose St Building<https://url.au.m.mimecastprotect.com/s/6O8_C1WLPxckoYOw1u4UkfVNAuX?domain=t…>) Zoom ID: https://uni-sydney.zoom.us/s/87306457455<https://url.au.m.mimecastprotect.com/s/JpqhC2xMQziRXw6ovixcxf5GEwn?domain=t…> Abstract: In this talk, we address data-driven output-feedback control of continuous-time linear systems using only measured input–output trajectories. The key idea is to employ Kreisselmeier’s adaptive filter to construct a canonical non-minimal realization of the unknown plant, thereby avoiding explicit system identification and full-state measurement. In the first part, we consider the stabilization of general MIMO linear systems from a finite input–output trajectory. By interpreting the adaptive filter as an observer for a stabilizable non-minimal realization, we formulate the controller synthesis problem as a data-dependent linear matrix inequality. A data-driven decomposition is introduced to separate the excited controllable component from the stable uncontrollable dynamics, thereby relaxing the required rank conditions. We also discuss conditions for informative data collection and a reduced-order filter implementation that improves computational scalability. In the second part, we briefly extend this framework to data-driven linear quadratic control. By augmenting the filter-generated realization, we develop a value-iteration method that learns an implementable output-feedback controller from input–output data while recovering the optimal state-feedback performance. Bio: Bowen Yi is an Assistant Professor in the Department of Electrical Engineering at Polytechnique Montréal, an engineering school affiliated with Université de Montréal, and a member of GERAD (Groupe d’études et de recherche en analyse des décisions). He received his Ph.D. degree from Shanghai Jiao Tong University, China, in 2019. From 2017 to 2019, he was a visiting researcher at the Laboratoire des Signaux et Systèmes (L2S), CNRS–CentraleSupélec, France. He subsequently held postdoctoral research positions at the Australian Centre for Robotics at the University of Sydney (2019 – 2022) and the Robotics Institute at the University of Technology Sydney (2022 – 2023). His research interests include nonlinear systems, particularly estimation, control, and learning, and their applications to robotics. He received the 2019 IEEE CCTA Best Student Paper Award and the Australian Research Council Discovery Early Career Researcher Award (ARC DECRA) in 2024. [https://d31hzlhk6di2h5.cloudfront.net/20260721/2e/8d/c7/1c/1c70baef8b350e38…] [Web Site]<https://url.au.m.mimecastprotect.com/s/fvjiC3QNPBiGlQxYwT5fPfQhmLQ?domain=t…> [LinkedIn]<https://url.au.m.mimecastprotect.com/s/kGp1C4QOPEimO16AyIqhDf4s-dA?domain=t…> [X]<https://url.au.m.mimecastprotect.com/s/A9SaC5QPXJix7GRYoiPiBfkxefy?domain=t…> [YouTube]<https://url.au.m.mimecastprotect.com/s/54pwC6XQ4Lf1vKMlEIwsGf5CApL?domain=t…> Copyright © 2026 The University of Sydney Please add acfr.admin(a)sydney.edu.au<mailto:acfr.admin@sydney.edu.au> to your address book or senders safe list to make sure you continue to see our emails in the future. Cricos Code 00026A TEQSA PRV12057 Disclaimer<https://url.au.m.mimecastprotect.com/s/ceXEC71R2NTQKkyjvhGtXfoLK5B?domain=t…> | Privacy statement<https://url.au.m.mimecastprotect.com/s/TDh6C81V0PTz0RQ8lhYukfyHfPt?domain=t…> | _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] Tour today 11am
by Jennifer Wakulicz 21 Jul '26

21 Jul '26

Hi everyone, Sorry for the late notice, but I'll be showing someone from Texas A&M University around the labs today at 11am. Please hide/showcase your work as appropriate! Cheers, Jen _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] J18 Level 2 Cafeteria
by Khalid Rafique 21 Jul '26

21 Jul '26

Hi J18 occupants Ag/Env has planned today a group lunch in the Level 2 Cafeteria between 12PM to 1.30PM. It would be great if you can use the place outside this time slot. Thank you for your understanding. Best regards, Khalid Rafique | CPEng, NER Program Manager Australian Centre for Robotics The University of Sydney Room 403, Building J18 | The University of Sydney | NSW | 2006 +61 2 8627 9070 | +61 2 9036 0000 (fax) | +61 427 670 296 khalid.rafique(a)sydney.edu.au<mailto:khalid.rafique@sydney.edu.au> | sydney.edu.au CRICOS 00026A This email plus any attachments to it are confidential. Any unauthorised use is strictly prohibited. If you receive this email in error, please delete it and any attachments. Please think of our environment and only print this email if necessary. _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] ACFR Seminar Series: Asst. Professor Bowen Yi - Input–Output Data-Driven Control of Continuous-Time Systems
by Sabrina Asri 21 Jul '26

21 Jul '26

Faculty of Engineering ACFR Seminar Series [https://d31hzlhk6di2h5.cloudfront.net/20260721/44/a0/ac/85/83de99ae1318da0a…] [https://d31hzlhk6di2h5.cloudfront.net/20260721/90/62/a1/a0/e806283be685b4d0…] Input–Output Data-Driven Control of Continuous-Time Systems Presented by Asst. Professor Bowen Yi Date: Wednesday, 22 July 2026, at 1:00 pm AEST Venue: ACFR seminar area, J04 level 2 (Rose St Building<https://url.au.m.mimecastprotect.com/s/6O8_C1WLPxckoYOw1u4UkfVNAuX?domain=t…>) Zoom ID: https://uni-sydney.zoom.us/s/87306457455<https://url.au.m.mimecastprotect.com/s/JpqhC2xMQziRXw6ovixcxf5GEwn?domain=t…> Abstract: In this talk, we address data-driven output-feedback control of continuous-time linear systems using only measured input–output trajectories. The key idea is to employ Kreisselmeier’s adaptive filter to construct a canonical non-minimal realization of the unknown plant, thereby avoiding explicit system identification and full-state measurement. In the first part, we consider the stabilization of general MIMO linear systems from a finite input–output trajectory. By interpreting the adaptive filter as an observer for a stabilizable non-minimal realization, we formulate the controller synthesis problem as a data-dependent linear matrix inequality. A data-driven decomposition is introduced to separate the excited controllable component from the stable uncontrollable dynamics, thereby relaxing the required rank conditions. We also discuss conditions for informative data collection and a reduced-order filter implementation that improves computational scalability. In the second part, we briefly extend this framework to data-driven linear quadratic control. By augmenting the filter-generated realization, we develop a value-iteration method that learns an implementable output-feedback controller from input–output data while recovering the optimal state-feedback performance. Bio: Bowen Yi is an Assistant Professor in the Department of Electrical Engineering at Polytechnique Montréal, an engineering school affiliated with Université de Montréal, and a member of GERAD (Groupe d’études et de recherche en analyse des décisions). He received his Ph.D. degree from Shanghai Jiao Tong University, China, in 2019. From 2017 to 2019, he was a visiting researcher at the Laboratoire des Signaux et Systèmes (L2S), CNRS–CentraleSupélec, France. He subsequently held postdoctoral research positions at the Australian Centre for Robotics at the University of Sydney (2019 – 2022) and the Robotics Institute at the University of Technology Sydney (2022 – 2023). His research interests include nonlinear systems, particularly estimation, control, and learning, and their applications to robotics. He received the 2019 IEEE CCTA Best Student Paper Award and the Australian Research Council Discovery Early Career Researcher Award (ARC DECRA) in 2024. [https://d31hzlhk6di2h5.cloudfront.net/20260721/2e/8d/c7/1c/1c70baef8b350e38…] [Web Site]<https://url.au.m.mimecastprotect.com/s/fvjiC3QNPBiGlQxYwT5fPfQhmLQ?domain=t…> [LinkedIn]<https://url.au.m.mimecastprotect.com/s/kGp1C4QOPEimO16AyIqhDf4s-dA?domain=t…> [X]<https://url.au.m.mimecastprotect.com/s/A9SaC5QPXJix7GRYoiPiBfkxefy?domain=t…> [YouTube]<https://url.au.m.mimecastprotect.com/s/54pwC6XQ4Lf1vKMlEIwsGf5CApL?domain=t…> Copyright © 2026 The University of Sydney Please add acfr.admin(a)sydney.edu.au<mailto:acfr.admin@sydney.edu.au> to your address book or senders safe list to make sure you continue to see our emails in the future. Cricos Code 00026A TEQSA PRV12057 Disclaimer<https://url.au.m.mimecastprotect.com/s/ceXEC71R2NTQKkyjvhGtXfoLK5B?domain=t…> | Privacy statement<https://url.au.m.mimecastprotect.com/s/TDh6C81V0PTz0RQ8lhYukfyHfPt?domain=t…> | _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] ACFR Seminar Series: Asst. Professor Jake Welde (Cornell) & Asst. Professor Thai Mai Thanh (VinUniversity)
by Sabrina Asri 21 Jul '26

21 Jul '26

Faculty of Engineering ACFR Seminar Series [https://d31hzlhk6di2h5.cloudfront.net/20260720/44/a0/ac/85/83de99ae1318da0a…] [https://d31hzlhk6di2h5.cloudfront.net/20260720/90/62/a1/a0/e806283be685b4d0…] 1) Symmetry and Beyond: Towards Efficient, Generalizable Learning for Low-Level Control of Aerial and Space Robots. 2) Advancing Healthcare Through Soft Robotics and Intelligent Systems Presented by Asst. Professor Jake Welde (Cornell) & Asst. Professor Thai Mai Thanh (VinUniversity) Date: Thursday, 23 July 2026, at 1:00 pm AEST Venue: ACFR seminar area, J04 level 2 (Rose St Building<https://url.au.m.mimecastprotect.com/s/BchzCGv0oyC0RPXLltLcPfB2L8G?domain=t…>) Zoom ID: https://uni-sydney.zoom.us/s/87306457455<https://url.au.m.mimecastprotect.com/s/L5lXCJyBrGf1v2MQzUWf9fyp_wV?domain=t…> **RSVP by COB Tuesday 21 July for catering purposes. Accept this calendar invitation if you plan to attend the seminar in person and join us for a light lunch at approximately 12.30pm. Abstract: Symmetry and Beyond: Towards Efficient, Generalizable Learning for Low-Level Control of Aerial and Space Robots To perform useful, interactive work, aerial and space robots must evolve beyond simple, single-body designs (e.g., quadrotors) towards high-dimensional, articulated morphologies. However, explicit analytical controllers for such systems are not easily crafted, and numerical optimal control algorithms quickly overwhelm the lightweight processors available onboard. Reinforcement learning frontloads these burdens to offline training, typically employing vast computational resources to run brittle training pipelines fine-tuned to particular tasks. To mitigate these downsides, we exploit the natural symmetries of robotic systems to efficiently learn high-performance, generalizable policies for low-level robot control. In particular, we develop a theory of symmetry reduction for tracking control problems, proving that a tracking controller trained in a reduced setting will perform equally well on the original system while also generalizing automatically to unseen trajectories. However, the classical notion of symmetry is too rigid to apply to many practical systems, which may enjoy only a very small symmetry group. We thus propose a relaxed notion of symmetry (termed “weak invariance”) that balances structure with generality, enabling us to factor out a much larger symmetry group while still attaining the same performance guarantees, greatly expanding the applicability and impact of these methods. Advancing Healthcare Through Soft Robotics and Intelligent Systems This talk presents recent advances in soft robotics and intelligent systems aimed at transforming healthcare and improving the quality of life. It highlights the development of flexible, bio-inspired robotic technologies for minimally invasive surgery, including soft robotic endoscopic platforms integrated with in situ 3D bioprinting and real-time navigation. The talk also explores wearable soft sensors and robotic devices for rehabilitation, enabling low-cost, accessible, and data-driven patient recovery. By combining soft materials, embedded systems, and artificial intelligence, these innovations offer safer, more adaptive, and patient-centered solutions. The presentation concludes with a vision for deploying such technologies in resource-constrained settings to address global healthcare challenges and enhance quality of life at scale. Bio: Jake Welde Jake Welde is an Assistant Professor with the Sibley School of Mechanical and Aerospace Engineering at Cornell University, where he leads the Geometry, Design, and Control Laboratory (or the “GeoDesiC Lab”). Prior to joining Cornell, he spent a decade at the University of Pennsylvania, where he completed his undergraduate and doctoral degrees in Mechanical Engineering and Applied Mechanics and his Masters in Robotics, working in the GRASP Laboratory with Vijay Kumar. He is broadly interested in how mathematical structure — symmetry, hierarchy, and mechanics — can be harnessed to create more capable algorithms for robot control, and how the interplay between control and morphology mediates overall robot capabilities. Thai Mai Thanh Dr. Thai Mai Thanh is an Assistant Professor of Mechanical Engineering at VinUniversity and Principal Investigator (PI) of the VinUni Biorobotics Lab. He earned his B.E. from Ho Chi Minh City University of Technology (2016), M.S. from KAIST (2019), and Ph.D. in Biomedical Engineering from UNSW Sydney (2023), where he also completed a postdoctoral fellowship in medical robotics. His research focuses on advancing healthcare through soft robotics, intelligent systems, and medical devices, with applications in minimally invasive surgery, in situ bioprinting, rehabilitation, and wearable haptics. He has authored over 40 publications and holds multiple international patents, with a strong emphasis on translating cutting-edge technologies into real-world clinical impact. [Register for Event >]<https://url.au.m.mimecastprotect.com/s/2Hb2CK1DvKTBNYprwSVhwf5deC4?domain=t…> [https://d31hzlhk6di2h5.cloudfront.net/20260720/2e/8d/c7/1c/1c70baef8b350e38…] [Web Site]<https://url.au.m.mimecastprotect.com/s/7EZ2CMwGxOtR1Lm68hAsKf80JsU?domain=t…> [LinkedIn]<https://url.au.m.mimecastprotect.com/s/bDhvCNLJyQUE8QyMxtWtDfyBiNH?domain=t…> [X]<https://url.au.m.mimecastprotect.com/s/Do5FCOMKzVT0jgQ2XtKuzfGWMil?domain=t…> [YouTube]<https://url.au.m.mimecastprotect.com/s/RkSmCP7LAXfoR2V5mtXC2fxbUD-?domain=t…> Copyright © 2026 The University of Sydney Please add acfr.admin(a)sydney.edu.au<mailto:acfr.admin@sydney.edu.au> to your address book or senders safe list to make sure you continue to see our emails in the future. Cricos Code 00026A TEQSA PRV12057 Disclaimer<https://url.au.m.mimecastprotect.com/s/BbjgCQnMBZf3pKZNqHjFmfGFzZ7?domain=t…> | Privacy statement<https://url.au.m.mimecastprotect.com/s/wu_GCROND2uR1pZglhMHxf1Jomn?domain=t…> | _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0

[Everybody at ACFR] ACFR Seminar Series: Asst. Professor Jake Welde (Cornell) & Asst. Professor Thai Mai Thanh (VinUniversity)
by Sabrina Asri 21 Jul '26

21 Jul '26

Faculty of Engineering ACFR Seminar Series [https://d31hzlhk6di2h5.cloudfront.net/20260720/44/a0/ac/85/83de99ae1318da0a…] [https://d31hzlhk6di2h5.cloudfront.net/20260720/90/62/a1/a0/e806283be685b4d0…] 1) Symmetry and Beyond: Towards Efficient, Generalizable Learning for Low-Level Control of Aerial and Space Robots. 2) Advancing Healthcare Through Soft Robotics and Intelligent Systems Presented by Asst. Professor Jake Welde (Cornell) & Asst. Professor Thai Mai Thanh (VinUniversity) Date: Thursday, 23 July 2026, at 10:00 am AEST Venue: ACFR seminar area, J04 level 2 (Rose St Building<https://url.au.m.mimecastprotect.com/s/BchzCGv0oyC0RPXLltLcPfB2L8G?domain=t…>) Zoom ID: https://uni-sydney.zoom.us/s/87306457455<https://url.au.m.mimecastprotect.com/s/L5lXCJyBrGf1v2MQzUWf9fyp_wV?domain=t…> **RSVP by COB Tuesday 21 July for catering purposes. Accept this calendar invitation if you plan to attend the seminar in person and join us for a light lunch at approximately 12.30pm. Abstract: Symmetry and Beyond: Towards Efficient, Generalizable Learning for Low-Level Control of Aerial and Space Robots To perform useful, interactive work, aerial and space robots must evolve beyond simple, single-body designs (e.g., quadrotors) towards high-dimensional, articulated morphologies. However, explicit analytical controllers for such systems are not easily crafted, and numerical optimal control algorithms quickly overwhelm the lightweight processors available onboard. Reinforcement learning frontloads these burdens to offline training, typically employing vast computational resources to run brittle training pipelines fine-tuned to particular tasks. To mitigate these downsides, we exploit the natural symmetries of robotic systems to efficiently learn high-performance, generalizable policies for low-level robot control. In particular, we develop a theory of symmetry reduction for tracking control problems, proving that a tracking controller trained in a reduced setting will perform equally well on the original system while also generalizing automatically to unseen trajectories. However, the classical notion of symmetry is too rigid to apply to many practical systems, which may enjoy only a very small symmetry group. We thus propose a relaxed notion of symmetry (termed “weak invariance”) that balances structure with generality, enabling us to factor out a much larger symmetry group while still attaining the same performance guarantees, greatly expanding the applicability and impact of these methods. Advancing Healthcare Through Soft Robotics and Intelligent Systems This talk presents recent advances in soft robotics and intelligent systems aimed at transforming healthcare and improving the quality of life. It highlights the development of flexible, bio-inspired robotic technologies for minimally invasive surgery, including soft robotic endoscopic platforms integrated with in situ 3D bioprinting and real-time navigation. The talk also explores wearable soft sensors and robotic devices for rehabilitation, enabling low-cost, accessible, and data-driven patient recovery. By combining soft materials, embedded systems, and artificial intelligence, these innovations offer safer, more adaptive, and patient-centered solutions. The presentation concludes with a vision for deploying such technologies in resource-constrained settings to address global healthcare challenges and enhance quality of life at scale. Bio: Jake Welde Jake Welde is an Assistant Professor with the Sibley School of Mechanical and Aerospace Engineering at Cornell University, where he leads the Geometry, Design, and Control Laboratory (or the “GeoDesiC Lab”). Prior to joining Cornell, he spent a decade at the University of Pennsylvania, where he completed his undergraduate and doctoral degrees in Mechanical Engineering and Applied Mechanics and his Masters in Robotics, working in the GRASP Laboratory with Vijay Kumar. He is broadly interested in how mathematical structure — symmetry, hierarchy, and mechanics — can be harnessed to create more capable algorithms for robot control, and how the interplay between control and morphology mediates overall robot capabilities. Thai Mai Thanh Dr. Thai Mai Thanh is an Assistant Professor of Mechanical Engineering at VinUniversity and Principal Investigator (PI) of the VinUni Biorobotics Lab. He earned his B.E. from Ho Chi Minh City University of Technology (2016), M.S. from KAIST (2019), and Ph.D. in Biomedical Engineering from UNSW Sydney (2023), where he also completed a postdoctoral fellowship in medical robotics. His research focuses on advancing healthcare through soft robotics, intelligent systems, and medical devices, with applications in minimally invasive surgery, in situ bioprinting, rehabilitation, and wearable haptics. He has authored over 40 publications and holds multiple international patents, with a strong emphasis on translating cutting-edge technologies into real-world clinical impact. [Register for Event >]<https://url.au.m.mimecastprotect.com/s/2Hb2CK1DvKTBNYprwSVhwf5deC4?domain=t…> [https://d31hzlhk6di2h5.cloudfront.net/20260720/2e/8d/c7/1c/1c70baef8b350e38…] [Web Site]<https://url.au.m.mimecastprotect.com/s/7EZ2CMwGxOtR1Lm68hAsKf80JsU?domain=t…> [LinkedIn]<https://url.au.m.mimecastprotect.com/s/bDhvCNLJyQUE8QyMxtWtDfyBiNH?domain=t…> [X]<https://url.au.m.mimecastprotect.com/s/Do5FCOMKzVT0jgQ2XtKuzfGWMil?domain=t…> [YouTube]<https://url.au.m.mimecastprotect.com/s/RkSmCP7LAXfoR2V5mtXC2fxbUD-?domain=t…> Copyright © 2026 The University of Sydney Please add acfr.admin(a)sydney.edu.au<mailto:acfr.admin@sydney.edu.au> to your address book or senders safe list to make sure you continue to see our emails in the future. Cricos Code 00026A TEQSA PRV12057 Disclaimer<https://url.au.m.mimecastprotect.com/s/BbjgCQnMBZf3pKZNqHjFmfGFzZ7?domain=t…> | Privacy statement<https://url.au.m.mimecastprotect.com/s/wu_GCROND2uR1pZglhMHxf1Jomn?domain=t…> | _______________________________________________ Everybody mailing list -- everybody(a)acfr.usyd.edu.au To unsubscribe send an email to everybody-leave(a)acfr.usyd.edu.au

1 0