To receive credit, Students must be either logged in with their waterloo account into zoom, or change their username formatted as follows prior to joining the Zoom call: First_Name Family_Name Student_ID Program (MASc/MEng/PhD), otherwise the attendance will not be counted. It is the student's responsibility to make sure that their username is properly formatted prior to joining Zoom, as changing it after joining will not be recorded with the correct format, the seminar organizers and the graduate advisor will not make adjustments after the end of the seminar.
Students must attend the full MME Department Research Seminars listed here and complete the Seminar Attendance Form.
To receive credit for attending the Seminar, please complete and submit this form.
You must login to view this form.
Degree Requirements
- MASC students are required to attend 8 seminars for degree completion
- MEng students are required to attend 4 seminars for degree completion
- Students must attend the entire seminar (at least 45 minutes) to receive credit
Note: To receive credit, students can only attend MME Department Research Seminars.
PhD Seminars | MASc Seminars | Past Seminars
PhD Seminars
PhD Comprehensive Examination
( Attendance is at the discretion of the supervisor and student who hold the exam)
Hamidreza Aghajani - Wednesday, June 5, 2024 from 9:00 AM to 12:00 PM Remote
Title: Additive Manufacturing of Non-Weldable Ni-Based Superalloys by Laser and Electron Beam Powder Bed Fusion Processes
Supervisor: Ehsan Toyserkani
Jigar Patel - Thursday, June 20, 2024, 11:00 AM to 2:00 PM E5 - 3006 in person
Title: Data-centric machine learning in laser powder bed fusion additive manufacturing
Supervisor: Mihaela Vlasea
Alireza Esfandbod - Thursday, July 4, 2024 from 9:30 AM-12:30 PM Remote
Title: A hierarchical agent-based model predictive control for vehicle applications
Supervisors: Amir Khajepour / Mohammad Pirani
Amirreza Koushki - Friday, July 26, 2024 from 12:00 PM to 3:00 PM Remote
Title: A Novel Perception-to-Motion Planning for Autonomous Driving using Imitation Learning
Supervisor: Amir Khajepour
Halil Ibrahim Yazici - Tuesday, August 6, 2024 from 9:00 AM to 12:00 PM E5 - 3052 Hybrid Format
Title: In situ optical diagnostics for aerosolized 2D material
Supervisors: Kyle Daun / Christof Schulz
MASc Seminars
Speaker: Dr. Jiajun Wu (Standford University)
Theme : machine learning
October 11th, 11AM-12PM.
(More information to follow)
Speaker: Prof. Wael Suleiman (University of Sherbrooke)
Theme: robotics
July 18th, 2-3PM, Online
(More information to follow)
Interaction of boundary layer instabilities and geometrical wall features: some recent work and ideas for next steps
Speaker: Marios Kotsonis, Prof. TU Delft
Theme: Fluids
--------------------------------------------
Time: 2024-July-3rd, 9:00 am-10:00 am EDT
Hybrid in E7-2343, Zoom meeting: (link)
Abstract: Boundary layer instabilities are of paramount importance for the dynamics of transitional flows, themselves dominating drag production and efficiency of many relevant systems such as aircraft wings.
However, in most real-life cases, aerodynamic surfaces are not smooth or geometrically perfect. Surface nonuniformities are inevitable due to debris, damages, and manufacturing features such as joints, panels, rivets etc. These geometrical features have a profound effect on boundary layer instabilities and transition. These effects are reviewed in this talk, based on recent experimental and numerical studies of our team. Expectedly, in most cases the effect of these features is detrimental, thus leading to anticipation of transition. However, we have discovered limited cases where roughness can actually lead to a delay of transition. These cases have revealed a wealth of fundamental interaction mechanisms, which can pave the way to passive flow control methods. Some next steps and ideas in moving forward will be shared.
Marios Kotsonis is Professor of Flow Control at Delft University of Technology, Faculty of Aerospace Engineering. His interests cover laminar-turbulent transition, flow instabilities, flow control and flow actuators, using theoretical, numerical and experimental techniques. He is recipient of the Veni and Vici grants of the Dutch Research Council and Starting, Proof-of-Concept and Consolidator grants of the
European Research Council.
Students must be either logged in with their waterloo account into zoom, or change their username formatted as follows prior to joining the Zoom call: First_Name Family_Name Student_ID Program(MASc/MEng/PhD), otherwise the attendance will not be counted. It is the student's responsibility to make sure that their username is properly formatted prior to joining Zoom, as changing it after joining will not be recorded with the correct format, the seminar organizers and the graduate advisor will not make adjustments after the end of the seminar.
Please contact the host, Prof. Yue Hu (yue.hu@uwaterloo.ca), if any questions
Versatile Imitation Learning and Reinforcement Learning with Motion Primitives in Robotics Speaker: Prof. Gerhard Neumann
Karlsruhe Institute of Technology, Germany
Theme: Robotics, Machine Learning
--------------------------------------------
Time: 2024-June-20th, 10:00-11:00 am EDT
Online, Zoom meeting: (link)
Abstract: In this talk, I will present our recent work on imitation learning and reinforcement learning in robotics. I will provide an overview of motion primitives (MPs) in robotics, which have been highly successful for smooth motion generation, and discuss their advantages and disadvantages compared to standard action-based policies. Additionally, I will introduce a novel MP representation called Probabilistic Dynamic Movement Primitives (ProDMPs). This representation can be easily integrated into a neural network architecture while ensuring smooth motion generation.
In the area of imitation learning, I will showcase new augmented reality-based user interfaces that facilitate low-effort data generation. Using these interfaces, we have collected diverse benchmark datasets for evaluating the latest imitation learning algorithms. Diffusion-based policies demonstrate impressive performance in this context. To enhance versatility, we extend the ProDMP framework with diffusion-based policies, ensuring smooth behavior representation.
For reinforcement learning (RL), I will highlight the beneficial exploration properties of MPs by exploring the parameter space of MPs rather than the robot's action space. I will introduce MP-based RL algorithms that treat the RL problem as a black-box optimization problem, requiring the algorithm to select MP parameters instead of low-level control actions. We have developed a novel RL algorithm incorporating differentiable trust region projection layers, which is well-suited for high-dimensional action spaces. This algorithm is extended for replanning and learning versatile behaviors.
Finally, I will discuss how approaches that incorporate human feedback can make reinforcement learning more practical for practitioners by eliminating the need for complex reward functions.
Prof. Gerhard Neumann is a full professor at the KIT and heading the chair "Autonomous Learning Robots" since Jan. 2020. Before that, he was group leader at the Bosch Center for AI and industry on campus professor at the University of Tübingen (from March to Dec. 2019) and full professor at the University of Lincoln in the UK (2016-2019). Gerhard completed my PhD in 2012 at the TU Graz and
was afterwards PostDoc and Assistant Professor at the TU Darmstadt. His research is focused on the intersection of machine learning, robotics and human-robot interaction. His goal is to create data-efficient machine learning algorithms that are suitable for complex robot domains, which includes learning from human feedback, versatile skill learning, learning learning to manipulate deformables, multiagent reinforcement learning as well as fundamental research in machine learning such as variational inference or meta-learning.
Students must be either logged in with their Waterloo account into zoom, or change their username formatted as follows prior to joining the Zoom call: First_Name Family_Name Student_ID Program(MASc/MEng/PhD), otherwise the attendance will not be counted. It is the student's responsibility to make sure that their username is properly formatted prior to joining Zoom, as changing it after joining will not be recorded with the correct format, the seminar organizers and the graduate advisor will not make adjustments after the end of the seminar.
Please contact the host, Prof. Yue Hu (yue.hu@uwaterloo.ca), if any questions
CONSTRAINED RADIAL BASIS FUNCTION (RBF) REGRESSION FOR THE MESHLESS AND BINLESS DATA ASSIMILATION IN IMAGE VELOCIMETRY
Speaker(s): Prof. M.A. Mendez and M. Ratz von Karman Institute, Belgium
Theme: Fluids and Solids
--------------------------------------------
Time: 2024-Apr-30th , 10:00 am-11:00 am EST
Zoom meeting: (link) Meeting ID: 789 699 0683
Passcode: MME2024
Abstract: Tracking-based image velocimetry is currently gaining prominence over traditional cross-correlation-based image velocimetry. Modern tracking methods can process images with high tracer density, generally enabling much higher resolution than cross-correlation-based velocimetry while circumventing its inevitable window or voxel averaging. On the other hand, tracking methods produce randomly scattered data, challenging the measurement post-processing to compute derivatives (e.g., velocity gradients or vorticity) or statistics (e.g., mean flow, turbulence intensity fields). A significant research effort in the community resulted in the development of many algorithms to map scattered velocity fields onto grids or bins, the first used to compute derivatives using traditional numerical methods (e.g. finite differences), the second to compute local statistics. Since these often combine data with physical priors (e.g. solenoidal field in incompressible flows), these algorithms fall in the data assimilation framework for image velocimetry. We recently proposed an approach for data assimilation in image velocimetry that requires no interpolation into grids or mapping into bins and achieves super-resolution. The approach uses constrained Radial Basis Functions (RBF) regression to derive an analytic expression of both the flow fields and their statistics. This talk presents the fundamentals of constrained RBF and showcases the method used in synthetic and experimental data to compute derivatives, flow statistics, and pressure fields. We also present our recently released open-source software package SPICY (Super-resolution and Pressure from Image veloCimetrY), developed at the von Karman Institute and close with a perspective on the current and future developments. The talk is based on these publications [1-3].
The talk will be given by Prof. M.A. Mendez and his student M. Ratz. M. A. Mendez is an associate professor at the von Karman Institute, leading the group on machine learning for applied fluid dynamics. His research group focuses on experimental and numerical methods, reduced-order modelling, flow control, digital twinning, and machine learning for industrial and environmental flows. M. Ratz is a PhD candidate at the von Karman Institute and at the Université Libre de Bruxelles, currently leading the developments of SPICY. His PhD focuses on data assimilation and numerical modelling of drone propeller’s aerodynamics.
References: [1] Sperotto, P., Pieraccini S., Mendez, M.A. (2022) A Meshless method to measure pressure fields from Image Velocimetry, Measurement Science and Technology, 33, 094005, Arxiv at https://arxiv.org/abs/2112.12752 [2] Sperotto, P., Ratz, M., Mendez, M.A (2023), SPICY: a Python toolbox for meshless assimilation from image velocimetry using radial basis functions, Journal of Open-Source Software 9(93), 5749, https://doi.org/10.21105/joss.05749 (open access) [3] Ratz, M., Mendez, M. A. (2024), A Meshless and binless approach to compute statistics in 3D Ensemble PTV, submitted to Experiments in Fluids. arvix at https://arxiv.org/abs/2403.11828.
Please contact the host, Prof. Zhao Pan (zhao.pan@uwaterloo.ca), if any questions.
Towards motor intelligence for soft robots
Speaker: Dr. Cosimo Della Santina
TU Delft, Netherlands
Theme: Robotics, Soft Robots
--------------------------------------------
Time: 2024-Apr-11th, 10:00-11:00 am EDT
Online, Zoom meeting: (link)
Abstract: Continuum soft robots are actuated mechanical systems whose body is composed of continuously deformable materials. In the past decade, substantial progress has been made in developing a low-level artificial brain for soft robotic systems that can make them execute precise motions and eventually exploit the intelligence embedded in their complex mechanical structures. In this talk, I will briefly introduce this grand challenge within the soft robotic field and then present
the model-based view of its solution. I will show how simplified models can be combined with nonlinear control theory and machine learning, leading to precise and dynamic task execution. I will conclude by presenting recent activities within my group concerning combining models with data and
transferring this body of knowledge toward the manipulation of soft objects.
Dr. Cosimo Della Santina is an Assistant Professor at TU Delft and a Guest Scientist at the German Aerospace Institute (DLR). He earned his Ph.D. in robotics (cum laude, 2019) from the University of Pisa. He was a visiting Ph.D. student and a postdoc (2017 to 2019) at MIT's Computer Science and Artificial Intelligence Laboratory. Subsequently, he held a senior postdoc position (2020) and served as a guest lecturer (2021) at the Department of Informatics at the Technical University of Munich (TUM). Cosimo has been awarded the uRobotics Georges Giralt Ph.D. Award (2020), the "Fabrizio Flacco" Young Author Award from I-RAS (2019), and was a finalist for the European Embedded Control Institute Ph.D. award (2020). In 2023, he received the IEEE AS Early Academic Career Award in Robotics and Automation. He is Principal Investigator for European and Dutch projects, such as H2020 Natural Intelligence, HE EMERGE, and Agrifood Nxtgen Hightech. He is an NWO ENI laureate and co-directs the Delft AI lab SELF. Cosimo leads the PhI-Lab at TU Delft, focusing on the study of embodied and disembodied intelligence in physical systems, with an emphasis on elastic and soft robots.
Please contact the host, Prof. Yue Hu (yue.hu@uwaterloo.ca), if any questions
Control of separated flow and suppression of flow-induced vibration with peripheral spinning rods
Speaker: Prof. Gustavo R. S. Assi
University of São Paulo, Brazil
Theme: Fluids and Thermal
--------------------------------------------
Time: Tue, April 9, 2024, 10 - 11 am EST
Location: In person in E7 2343 or Zoom (link) Meeting ID: 789 699 0683
Passcode: MME2024
Abstract: Rotating surfaces are an interesting tool for controlling flow separation around bluff bodies. When spinning rods are strategically placed around a circular cylinder, they disrupt the flow in the boundary layer and near wake, effectively delaying separation and vortex formation. This technique not only influences flow patterns but also reduces separation and mitigates vibration effects induced by the flow, thus helping to minimize vortex-induced vibration. With sufficient rotation, a system of rotating actuators can even neutralize drag or generate beneficial lateral forces. In this presentation, we will discuss experimental and numerical findings from our investigation involving eight rotating rods positioned around a central circular cylinder. Our focus will be on unravelling the fundamental physical mechanisms at play. This innovative approach holds great potential for diverse engineering applications, especially in the oceanic, aerospace, automotive, and civil engineering sectors, where managing flow separation, vibration, and noise is important for optimal performance and safety. Of particular interest, this active flow-control method can be applied to large ocean systems, such as floating offshore platforms with cylindrical hulls, offering new ideas for enhancing their efficiency and stability.
Dr. Gustavo R. S. Assi is the Professor of Renewable Energy and Energy Transition Technologies in the Department of Naval Architecture and Ocean Engineering at the University of São Paulo, Brazil. Deputy Director of the OTIC Offshore Technology Innovation Centre. Member of the Specialist Committee on Ocean Renewable Energy of the ITTC International Towing Tank Conference. A Naval Engineer from USP and PhD in Aeronautics from Imperial College London, he has worked as a Visiting Scholar at the University of Oxford (2013-14), Caltech (2016-17), and is currently at Texas A&M University (2023-24). His research interests include fluid-structure interaction, flow-induced vibration, experimental and computational fluid dynamics, ocean renewable energy, engineering education, science communication, and philosophy of technology.
Please contact the host, Profs. Serhiy Yarusevych (syarus@uwaterloo.ca, in person) and Zhao Pan
(zhao.pan@uwaterloo.ca, online), if any questions.
Let droplets drop the temperature: Enhancing heat transfer using droplets
Speaker: Patricia Weisensee, PhD
Washington University at St. Louis, USA
Theme: Fluids and Thermal
--------------------------------------------
Time: 2024-Feb-26th, 2:00 pm-3:00 pm EST
Online, Zoom meeting: (link) Meeting ID: 789 699 0683
Passcode: MME2024
Abstract: We experience and rely on droplets almost every single day of our lives: some more familiar (the shower in the morning, the rain on our way to work, printing documents once at work, etc.), some less obvious (atmospheric water harvesting, thermal management & power generation, materials manufacturing and processing). Yet we hardly ever think about them – we take them for granted. In this talk I will show that droplets aren’t only ubiquitous in both nature and industrial processes, but also a complex and fascinating research subject that still holds many mysteries to be solved. In my research group, we are especially interested in the coupling of fluid dynamics, heat transfer, and phase change (condensation or evaporation) during the interaction of droplets with solid or other liquid surfaces.
In this presentation, I will introduce two examples of such interactions: 1) droplet impact and evaporation/boiling dynamics on heated surfaces, and 2) dropwise condensation on so-called lubricant-infused surfaces. These oil-coated surfaces can significantly increase water collection rates compared to bare metal surfaces due to extremely high droplet mobility. Lubricant wetting ridges surrounding droplets introduce an attractive capillary force, leading to self-propelled and gravity-independent droplet motion, which efficiently clears the surface for frequent re-nucleation. On the other hand, restricting the mobility of droplets can be advantageous during droplet evaporation. Interestingly, when droplets impact a heated surface, the creation of additional contact lines, either through wettability-patterning the surface or the formation of an entrapped air bubble, does not significantly alter the heat transfer performance. Instead, convection (at early times) and conduction (at later times) dominate heat transfer, meaning that the addition of – for example – metallic posts with high thermal conductivity on the surface can effectively increase heat transfer rates in these scenarios.
Dr. Weisensee is an Assistant Professor in the Department of Mechanical Engineering & Materials Science at Washington University in St. Louis (WashU). She earned her PhD in Mechanical Engineering from the University of Illinois at Urbana-Champaign (UIUC) in 2016. She received a Diplom-Ingenieur in Mechanical Engineering from TU Munich in 2013 and also holds a M.S. in Materials Sciences from UIUC (2011). At WashU, Dr. Weisensee leads the Thermal Fluids Research Group, which focuses on understanding the interplay of fluid dynamics and heat transfer of droplets and other multi-phase systems, with applications in thermal management, water harvesting, additive manufacturing, and droplet interactions with natural and engineered systems. To fundamentally study these thermal-fluidic interactions, her group combines multiple experimental techniques, such as high-speed optical and infrared (IR) imaging, interferometry, confocal fluorescence microscopy, and conventional heat transfer measurements. Dr. Weisensee is a recipient of the NSF CAREER Award, the NASA Early Career Faculty Award, the AFOSR YIP Award, the 2014 Siemens Energy Award, the 2020 ASME ICNMM Outstanding Early Investigator Award, as well as the St. Louis-wide 2020 Emerson Excellence in Teaching Award.
Please contact the host, Prof. Zhao Pan (zhao.pan@uwaterloo.ca), if any questions
Machine Learning and Control for Robot Interaction Autonomy
Speaker: Dr. Arash Ajoudani
Human-Robot Interfaces and Interaction (HRI2), Italian
Institute of Technology, Italy
Theme: Robotics, Machine Learning
--------------------------------------------
Time: 2024-Mar-6th, 10:00-11:00 am EST
Online, Zoom meeting: (link)
Abstract: Future robots are expected to help us in various tasks, of which many require collaborative effort to be successfully completed. To achieve such behaviours, the robot must be able to physically interact with the human counterpart and predict the intentions of the human counterparts, despite the additional interaction with unpredictable environments. Appropriate robot control methods therefore are essential for the robot to deal with various uncertainties, dynamic aspects, and complex task requirements. A promising direction towards this goal is to adaptively regulate
mechanical parameters of the robot via software, based on the task requirements and robot proprioceptive and exteroceptive sensory information. This talk will explore how the Human-Robot Interfaces and Interaction (HRI²) laboratory is using machine learning and control to create more autonomous yet truly collaborative robots, ultimately leading to more adaptable, more efficient interactions in various settings. The talk will cover topics related to real-time human kinodynamic state monitoring in HRI/C and autonomous robot loco-manipulation controllers using hybrid ML and control.
Dr. Arash Ajoudani is the director of the Human-Robot Interfaces and Interaction (HRI²) laboratory at IIT. He is a recipient of the European Research Council (ERC) proof-of-concept grant 2023 Real-Move and the ERC starting grant 2019 (Ergo-Lean), the coordinator of the Horizon-2020 project SOPHIA, the co-coordinator of the Horizon-2020 project CONCERT, and a principal investigator of the HORIZON-MSCA project RAICAM, and the national projects LABORIUS and COROMAN. He is a recipient of the IEEE Robotics and Automation Society (RAS) Early Career Award 2021, and winner of the SmartCup Liguria award 2023, Amazon Research Awards 2019, of the Solution Award 2019 (MECSPE2019), of the KUKA Innovation Award 2018, of the WeRob best poster award 2018, and of the best student paper award at ROBIO 2013. His PhD thesis was a finalist for the Georges Giralt PhD award 2015 - best European PhD thesis in robotics. He was also a finalist for the best paper award on mobile manipulation at IROS 2022, the best paper award at Humanoids 2022 (oral category), the Solution Award 2020 (MECSPE2020), the best conference
paper award at Humanoids 2018, the best interactive paper award at Humanoids 2016, the best oral presentation award at Automatica (SIDRA) 2014, and for the best manipulation paper award at ICRA 2012.
He is the author of the book "Transferring Human Impedance Regulation Skills to Robots" in the Springer Tracts in Advanced Robotics (STAR), and several publications in journals, international conferences, and book chapters. He is currently serving as an elected IEEE RAS AdCom member (2022-2024), and as chair and representative of the IEEERAS Young Professionals Committee, and as a Senior Editor of the International Journal of Robotics Research (IJRR).
He has been serving as a member of scientific advisory committee and as an associate editor for several international journals and conferences such as IEEE RAL, ICRA, IROS, ICORR, etc. He is a scholar of the European Lab for Learning and Intelligent Systems (ELLIS). His main research interests are in physical human-robot interaction, mobile manipulation, robust and adaptive control, assistive robotics, and tele-robotics.
Please contact the host, Prof. Yue Hu (yue.hu@uwaterloo.ca), if any questions.