4th of June 2021 (Time Zone GMT-04)
08:30-08:40 Welcome [14:30-14:40 CEST]
Session-1:
08:40-09:00 Jaeheung Park, Seoul National University [14:40-15:00 CEST] video
Topic: Towards compliant and precise control of legged robots for walking
More real-world applications of legged robots are being considered due to recent advances in research. Especially, four-legged robots have progressed so much that they can operate in various terrains. Two-legged robots, however, still have more difficulties in terms of the capability of dynamic motion and walking over uneven terrain. Among many aspects, I would like to introduce recent efforts in our research group. One is to estimate impact or disturbance to the robot, and use the information to create compliant motion for robust walking. This features non-use of force/torque sensor or joint torque sensors. Another direction is to improve position tracking performance of torque-controlled robots, which is a common difficulty when using torque controlled robots. Various hardware experiments will be demonstrated as examples.
BioJaeheung Park is a Professor in the Department of Intelligence and Information at Seoul National University, South Korea, since 2009. Prior to joining Seoul National University, he shortly worked at Hansen Medical Inc, a medical robotics company. He received the Ph.D. degree from Stanford University, and the B.S. and M.S. degrees from Seoul National University. His research group is currently conducting many national projects on the topics of humanoid robots, rehabilitation/medical robots, and autonomous vehicles. He was a team leader of TEAM SNU for DRC Finals 2015 (DARPA Robotics Challenge Finals), which was a robotics competition for disaster response. In 2016, His team won the award given by Minister of Science, ICT, and Future Planning, at Challenge Parade that was held by the Korean Government. He has participated in organizing many international robotics conferences such as IROS, HUMANOIDS, and HRI. He has also served as a Conference Editorial Board or as an Associate Editor for many international conferences. He currently serves as a co-chair of the RAS Technical Committee on Whole-body Control.
09:00-09:20 Sangbae Kim, MIT [15:00-15:20 CEST]
Topic: Legged robots with Proprioceptive actuation
This talk introduces recent research activities on MIT Minicheetahs regarding navigation and jumping controllers. In addition, this talk will introduce a newly developed MIT humanoid robot and a preliminary controller.
BioProf. Sangbae Kim, is the director of the Biomimetic Robotics Laboratory and a Professor of Mechanical Engineering at MIT. His research focuses on the bio-inspired robot design by extracting principles from animals. Kim’s achievements on bio-inspired robot development include the world’s first directional adhesive inspired from gecko lizards, and a climbing robot, Stickybot, that utilizes the directional adhesives to climb smooth surfaces featured in TIME’s best inventions in 2006. Recent achievement includes the development of the MIT Cheetah capable of stable outdoor running up to 13mph and autonomous jumping over an obstacles at an efficiency of animals. This achievement was covered by more than 300 media articles. He is a recipient of best paper award from International Conference on Robotics and Automation (2007), King-Sun Fu Memorial Transactions on Robotics (2008) and IEEE/ASME transactions on mechatronics (2016), DARPA Young Faculty Award (2013), NSF CAREER award (2014), and Ruth and Joel Spira Award for Distinguished Teaching (2015).
09:20-09:40 Jessy Grizzle & Maani Ghaffari, University of Michigan [15:20-15:40 CEST]
Topic: Legged Robot Autonomy
We will report on autonomous navigation and semantic mapping experiments conducted with the Cassie Blue bipedal robot and the MIT mini-Cheetah robot, both at the University of Michigan. The sensors attached to Cassie include an IMU, a 32-beam LiDAR and an RGB-D camera; for min-Cheetah, an IMU and RGB-D camera. The talk will discuss the integration of perception, sensor fusion and calibration, map building, planning, and low-level feedback control. We conclude by discussing challenges that have emerged from our field tests in search of real-world autonomy for legged robots. Most of our software has been made open-source; that which is not yet released will be soon.
Bio (Jessy) Jessy W. Grizzle is a (full) Professor in the Control Systems Laboratory of the Electrical Engineering and Computer Science Department within the College of Engineering at The University of Michigan in Ann Arbor, Michigan. He is the Director of Michigan Robotics and a member of the ECE Systems Laboratory and the Robotics and Computer Vision Area. He has a courtesy appointment in Mechancial Engineering. My primary research area used to be the theory of nonlinear control systems. While I still have strong interest in this subject and feel a sense of community with that body of researchers, my research activities have significantly broadened over time. My work now covers the control of bipedal robots and correct-by-construction control methods for Advanced Driver Assist Systems. From 1986 to 2010, I worked on various aspects of modeling and control of automotive powertrain systems and control of HEVs. From 1991-2001, I applied systems and control techniques to improve the operation of plasma-based microelectronics manufacturing equipment.
Bio (Maani)Maani Ghaffari is an Assistant Professor in Naval Architecture and Marine Engineering and Robotics at The University of Michigan in Ann Arbor. He received the Ph.D. degree from the Centre for Autonomous Systems (CAS), University of Technology Sydney, NSW, Australia, in 2017. He recently established the Computational Autonomy and Robotics Laboratory. His research interests lie in the theory and applications of robotics and autonomous systems.
09:40-10:00 Wang Xingxing, Unitree Robotics [15:40-16:00 CEST] video
Topic: Breakthrough in the marketization of quadruped robots
Since establishment in 2016, Unitree Robotics has been committed to promoting the marketization of quadruped robots since. We have been thinking about how to promote the growth more quickly. At present, based on the current four-legged robot technology level, manufacturing costs, product functions, product advantages, retail prices and other factors, we hope to promote greater market growth.
BioWang Xingxing is the founder, CEO&CTO of Unitree Robotics. During his master’s degree (year 2013 to 2016), he independently completed the high-performance low-cost quadruped robot XDog, which is driven by the outer rotor brushless motor.Established Unitree Robotics at the end of 2016, continued to promote the global marketization of high-performance low-cost quadruped robots.Unitree Robotics has successively launched LaikaGo, LaikaGo Pro, AlienGo and A1, which are widely recognized by the market.
10:00-10:20 Avik De, Ghost Robotics [16:00-16:20 CEST]
Topic: Perception-aided control and planning with reduced order models
Some benefits of utilizing reduced-order “template” models for control are some degree of robustness and insensitivity to variations in system parameters, as well as reduction in complexity and computational burden. While the workflow for implementing reactive or optimization-based control strategies that act on template models is well-understood, it is not yet obvious how best to incorporate perception information into this paradigm.
BioAvik is co-founder and CTO of Ghost Robotics, a startup company commercializing legged robotics in Philadelphia. The company currently has two products, a 45kg quadruped aimed at industrial applications, and a 12kg quadruped aimed at research applications. Previously, Avik completed a postdoc at Harvard SEAS advised by Rob Wood, where he researched design of micro-scale flapping robots, as well as strategies for their control. He received his PhD in Sep 2017, at the GRASP laboratory (Kodlab) in the University of Pennsylvania advised by Dan Koditschek. The main thread tying all of his work has been bio-inspired design and control strongly anchored in empirical robotics. His research has focused on examining the strengths and weaknesses of modular and hierarchical control strategies, as well as demonstrating efficient and effective control of dynamic locomotion in a way that generalizes across platforms (quadruped, tailed biped, …) and behaviors (hopping, running, …).
10:20-10:50 Coffee break [16:20-16:50 CEST]
Session-2:
10:50-11:10 Claudio Semini, IIT [16:50-17:10 CEST]
Topic: Novel Methods for Quadruped Locomotion over Rough Terrain
In this talk I will present some of the most recent quadruped locomotion control and planning results achieved by IIT’s Dynamic Legged Systems lab. First, I’ll show how our previously introduced Feasible Region can be extended to include kinematic limits and dynamic motions. Then I’ll present our nonlinear MPC with mobility optimization for quadruped locomotion over uneven terrain. Next, I will introduce ViTAL, which is an online vision-based locomotion planning strategy that selects the footholds based on the robot capabilities (skills), and plans the robot pose that maximizes the chances of the robot succeeding in reaching these footholds. Last, I’ll give an update about our most recent work on Tele-manipulation with HyQReal.
BioClaudio Semini is the head of IIT’s Dynamic Legged Systems (DLS) lab. He holds an MSc degree from ETH Zurich in electrical engineering and information technology. He spent 2 years in Tokyo for his research: MSc thesis at the Hirose Lab at Tokyo Tech and staff engineer at the Toshiba R&D center in Kawasaki working on mobile service robotics. During his PhD and subsequent PostDoc at IIT, he developed the quadruped robot HyQ and worked on its control. Since 2012 he leads the DLS lab. His research focus lies on the construction and control of highly dynamic and versatile legged robots in real-world environments. He has published around 100 articles in journals and peer-reviewed conferences and received several awards for them. He is a co-founder and chair of the IEEE RAS Technical Committee on Robot Mechanisms and Design. He was the coordinator of the EU-funded ECHORD++ experiment HyQ-REAL (www.hyq-real.eu) and currently heads the IIT side of the Moog/IIT joint lab.
11:10-11:30 Oliver Stasse, LAAS-CNRS [17:10-17:30 CEST]
Topic: Memory of Motion: An approach to combine efficiently Model Predictive Control and Neural Networks
What if we could generate complex movements for arbitrary robots with arms and legs interacting in a dynamic environment in real-time? Such a technology would certainly revolutionize the motion capabilities of robots and unlock a wide range of very concrete industrial and service applications: robots would be able to react in real-time to any change of the environment or unexpected disturbance during locomotion or manipulation tasks. However, the computation of complex movements for robots with arms and legs in multi-contact scenarios in unstructured environments is not realistically amenable to real-time with current computational capabilities and numerical algorithms. The project Memmo aims to solve this problem by 1) relying on massive off-line caching of pre-computed optimal motions that are 2) recovered and adapted online to new situations with real-time tractable model predictive control and where 3) all available sensor modalities are exploited for feedback control going beyond the mere state of the robot for more robust behaviors. We will show the current results we have been able to achieve on the humanoid robot TALOS.
BioOlivier Stasse is a CNRS Senior scientist in humanoid robotics at LAAS-CNRS, Toulouse France. His research interest is in fast decision making to generate motion for humanoid robotics. He managed the specifications and the contracting of the first humanoid robot of the TALOS serie from PAL-ROBOTICS. He is co-directing the joint lab ROB4FAM with Airbus Toulouse, and is currently serving as an Associate Editor for the IEEE Transactions on Robotics. He received in 2000 a Ph.D. on Intelligent Systems from the University of Paris 6, and the French Habilitation to Supervise Research (HDR) in Robotics (2013) from the University of Toulouse III. From 2000 to 2003, he was assistant professor at the Univ. of Paris XIII. From 2003 to 2011, he was at the Joint French-Japanese Robotics Laboratory (JRL) between the CNRS and the AIST in Tsukuba. In 2011 he joined the Gepetto team at LAAS, and is leading the team since 2021.
11:30-11:50 Auke Ijspeert, EPFL [17:30-17:50 CEST]
Topic: Investigating animal legged locomotion using biorobots
The ability to efficiently move in complex environments is a fundamental property both for animals and for robots, and the problem of locomotion and movement control is an area in which neuroscience, biomechanics, and robotics can fruitfully interact. In this talk, I will present how biorobots and numerical models can be used to explore the interplay of the four main components underlying animal locomotion, namely central pattern generators (CPGs), reflexes, descending modulation, and the musculoskeletal system. Going from salamander to human locomotion, I will present a series of models that tend to show that the respective roles of these components have changed during evolution with a dominant role of CPGs in salamander locomotion, and a more important role for sensory feedback and descending modulation in mammalian and human locomotion. Interesting properties for legged robot and lower-limb exoskeleton locomotion control will finally be discussed.
BioAuke Ijspeert is a professor at EPFL (Lausanne, Switzerland) since 2002, and head of the Biorobotics Laboratory. He has a BSc/MSc in physics from EPFL (1995), a PhD in artificial intelligence from the University of Edinburgh (1999). He is an IEEE Fellow. His research interests are at the intersection between robotics, computational neuroscience, nonlinear dynamical systems and applied machine learning. He is interested in using numerical simulations and robots to gain a better understanding of animal locomotion, and in using inspiration from biology to design novel types of robots and controllers. He is also investigating how to assist persons with limited mobility using exoskeletons and assistive furniture.
11:50-12:10 Luca Marchioni, PAL Robotics [17:50-18:10 CEST]
Topic: Making of a jumping robot
PAL robotics has been developing technology for a new bipedal platform that will enable research in dynamic locomotion and advance towards more capable and robust humanoid robots. The project is named Kangaroo as the robot has been designed thinking in highly dynamic motions such as jumps. The talk will cover the hardware design and an overview of the control system of the robot as well as presents preliminary experiments with prototype actuators for the leg length and hip joints.
BioLuca Marchionni obtained his BSc in Control Engineering in 2005 and his MSc in Systems and Control Engineering in 2007, both at the University of Rome La Sapienza. He joined PAL Robotics in 2009 as Software and Control Engineer, where he participated in the development of wheeled and biped robots, taking care of dynamic simulations, autonomous navigation and bipedal walking software. In 2014, Luca became the CTO, in charge of supervising research and development activities and ensuring that they serve the company’s overall business strategy. Currently, he is also the Project Manager of the TALOS robot.
12:10-12:30 Jonathan Hurst, Agility Robotics [18:10-18:30 CEST]
Topic: Robots for a Human World
Legged locomotion is cyclic dynamical behavior, but legged robots in the real world must do much more than just locomote. Still, the lowest-level dynamics, even the passive dynamics, are a critical foundation for everything: and levels of a control hierarchy are emergent from features of the behaviors. For example, impact dynamics are governed by inertias and compliances in a mechanism; but control of a gait is governed by step placement. We are working our way up from the bottom, and in this talk, I show a number of examples of behaviors through different levels of the control hierarchy. I’ll end with another important feature of legged robot deployments in the real world: our business vision at Agility Robotics.
BioJonathan W. Hurst is a Professor of Robotics, co-founder of the Oregon State University Robotics Institute, and Chief Technology Officer and co-founder of Agility Robotics. He holds a B.S. in mechanical engineering and an M.S. and Ph.D. in robotics, all from Carnegie Mellon University. His university research focuses on understanding the fundamental science and engineering best practices for legged locomotion. Investigations range from numerical studies and analysis of animal data, to simulation studies of theoretical models, to designing, constructing, and experimenting with legged robots for walking and running, and more recently, using machine learning techniques merged with more traditional control to enable highly dynamic gaits. Agility Robotics is extending this research to commercial applications for robotic legged mobility, working towards a day when robots can go where people go, generate greater productivity across the economy, and improve quality of life for all.
12:30-13:20 Extended Abstract (Poster) Session [18:30-19:20 CEST]
Session-3:
13:20-13:50 Marco Hutter & Péter Fankhauser, ETH & ANYbotics [19:20-19:50 CEST]
Topic: Tough anymals
In this talk we will give some insights what is needed to make the legged robot ANYmal a fully certified and field-ready quadrupedal robot. We present several use-cases and real-world deployments from industrial inspection to autonomous exploration of underground areas.
Bio (Marco)Marco is an assistant professor for robotic systems at ETH Zurich and co-founder of ANYbotics AG, a Zurich-based company developing legged robots for industrial applications. Marco’s research interests are in the development of novel machines and actuation concepts together with the underlying control, planning, and learning algorithms for locomotion and manipulation. His works find application from electrically actuated quadrupeds like ANYmal to large-scale autonomous excavators used for digital fabrication and disaster mitigation. Marco is part of the National Centre of Competence in Research (NCCR) Robotics and NCCR Digital Fabrication and PI in various international projects (e.g. EU Thing, NI) and challenges (e.g. DARPA SubT).
Bio (Peter)Péter Fankhauser is co-founder and CEO of the Zurich based startup company ANYbotics, developing autonomous mobile robots for harsh environments such as the legged robot ANYmal. In his previous role, he led the company’s business development and marketing activities. During his doctoral degree at ETH Zurich, Péter developed mapping, planning, and control strategies for legged robots to climb over rough terrain.
13:50-14:10 Marco Da Silva, Boston Dynamics [19:50-20:10 CEST]
Topic: Applying Spot to Real World Applications
Spot, Boston Dynamics quadrupedal robot product, can reliably walk nearly anywhere a human can, but what does it do? While legged mobility is a necessary skill for many use cases, it’s not the only requirement for valuable producing applications. This talk will focus on the work we have done to develop Spot for remote and autonomous sensing applications: i) Encapsulating Spot’s mobility in an extensible API, ii) Building an autonomous capability, iii) Making it easy to add sensing to build value producing solutions.
BioDr. Marco da Silva is a Principal Roboticist at Boston Dynamics. Marco currently manages the engineering team on Spot that develops autonomy, payloads and other software to make Spot useful for industrial applications. Prior to Spot, he was Atlas Platform Director at Boston Dynamics, a program to develop the world’s most advanced biped robot. Marco joined Boston Dynamics in 2010 after graduating with a PhD from MIT. Prior to that Marco worked at Pixar Animation Studios from 2001 to 2005.
14:10-14:30 Daniel Koditschek, University of Pennsylvania [20:10-20:30 CEST]
Topic:
A Palette of Sagittal Plane Template Compositions Toward Reactive Transitional Legged Robot Behaviors
AbstractThe talk will present working implementations on a quadrupedal robot of highly energetic increasingly reactive traversal through complicated substrates with highly varied geometry and friction. We will outline the synthesis procedure used to guarantee that the parallel compositions anchor the designated templates. We will briefly mention some of the key challenges to the overall agenda including: the formal problem of representing anchor obstacles and goals in the template state space as well as the empirical problem of quickly and accurately sensing the instantaneous hybrid mode. click here for the full abstract.
Bio (Daniel)Daniel E. Koditschek is the Alfred Fitler Moore Professor of Electrical and Systems Engineering, within the University of Pennsylvania School of Engineering and Applied Science. Koditschek received his bachelor’s degree in Engineering and Applied Science and his M.S. and Ph.D. degrees in Electrical Engineering in 1981 and 1983, all from Yale University. He served on the Yale Faculty in Electrical Engineering until moving to the University of Michigan a decade later. In January 2005, he moved to Penn as Chair of the recently formed Electrical and Systems Engineering Department, a position which he held through 2012. Koditschek is a member of the AMS, ACM, MAA, SIAM, SICB and Sigma Xi and is a Fellow of the IEEE and the AAAS. He holds a 2016 US Vannevar Bush Faculty Fellowship awarded by the US Assistant Secretary for Defense (Research and Engineering). He was awarded the 2016 IEEE Robotics and Automation Society Pioneer Award and the 2017 Penn Heilmeier Award for Excellence in Research.
Bio (Turner)T. Turner Topping received his B.S. and M.S. in Electrial Engineering from the University of Southern California in 2014. He is currently pursuing a PhD under the direction of Daniel Koditschek, studying the design and composability of legged locomotion behaviors within KodLab, part of the larger GRASP lab at the University of Pennsylvania.
14:30-14:45 Feifei Qian, USC [20:30-20:45 CEST]
Topic: Generating desired terrain responses to achieve effective locomotion from complex environments
Bringing legged robots to real-world operation requires effective locomotion over non-flat, non-rigid terrains, such as sand, debris, and rubble field. In this talk, I will present our efforts in addressing this challenge using a Robophysics approach — integrating robotics with terradynamics and locomotion biomechanics to create reduced-order interaction models that can enable prediction of terrain responses and resulting locomotor performance. I will show that these simplified models can already allow low-cost legged robots to elicit desired terrain responses to produce effective locomotion on granular and obstacle-cluttered substrates. In our on-going work, with the advancement of direct-drive actuator technology, we are developing desert-exploring robot that can use individual leg as soil strength sensor to monitor terrain responses. These terrain response measurements, integrated with our interaction models and representations, are leading to novel strategies for environment-aided robot locomotion and navigation, better understandings of animal gait transition behaviors, and new decision-making workflows for human-robot collaborative environment explorations.
BioFeifei Qian is an Assistant Professor of Electrical and Computer Engineering at University of Southern California. Qian received her PhD in Electrical Engineering and M.S. in Physics from Georgia Institute of Technology, in 2015 and 2011, respectively. Prior to her appointment at USC, she worked in the GRASP lab at University of Pennsylvania as a postdoctoral fellow. Her expertise is in analyzing and modeling the complex interactions between robots and environments, and developing innovative control and sensing strategies to improve robot mobility on challenging terrains. Her work was awarded the best student paper of Robotics: Science & Systems and has been featured in BBC News, Phys.org, and R&D Magazine.
14:45-15:00 Hsiu-Chin Lin, McGill University [20:45-21:00 CEST]
Topic: A generic model-based control framework in unstructured environments
One key aspect of integrating robots into unstructured environments is that robots need to respond and adapt to unexpected disturbances and various environment. In this talk, I will summarize some of my work on optimizing the controller with respect to external disturbances. I will also discuss my work on the estimation of movement constraints imposed by the tasks or the environment in order to adopt robot motion. The applications were verified on both manipulation and locomotion on real robotic platforms.
BioHsiu-Chin Lin is an Assistant Professor in the Computer Science and the Department of Electrical and Computer Engineering at McGill University. Her research spans model-based motion control, optimization, and machine learning for motion planning. She is particularly interested in adapting robot motion in dynamic environments for manipulators and quadruped robots. Prior to McGill, she was a Research Associate at the University of Edinburgh and the University of Birmingham. She received her Ph.D. from the University of Edinburgh for her work on robot learning.
15:00-15:15 Koushil Sreenath, UC Berkeley [21:00-21:15 CEST]
Topic: Dynamic Loco-Manipulation
In this talk I will summarize our recent results on combining locomotion and manipulation using legged robots. The idea is to use legs not just for locomotion but also for manipulation. We will see how bipedal and quadrupedal robots can be used for manipulating a ball, a payload, and even a blind-folded human. I will present simulation and experimental results demonstrating loco-manipulation.
BioKoushil Sreenath is an Assistant Professor of Mechanical Engineering, at UC Berkeley. He received a Ph.D. degree in Electrical Engineering and Computer Science and a M.S. degree in Applied Mathematics from the University of Michigan at Ann Arbor, MI, in 2011. He was a Postdoctoral Scholar at the GRASP Lab at University of Pennsylvania from 2011 to 2013 and an Assistant Professor at Carnegie Mellon University from 2013 to 2017. His research interest lies at the intersection of highly dynamic robotics and applied nonlinear control. His work on dynamic legged locomotion on the bipedal robot MABEL was featured on The Discovery Channel, CNN, ESPN, FOX, and CBS. His work on dynamic aerial manipulation was featured on the IEEE Spectrum, New Scientist, and Huffington Post. His work on adaptive sampling with mobile sensor networks was published as a book entitled Adaptive Sampling with Mobile WSN (IET). He received the Best Paper Award at the Robotics: Science and Systems (RSS) Conference in 2013, and the Google Faculty Research Award in Robotics in 2015.
15:15-15:30 Ryan Steindl, CSIRO [21:15-21:30 CEST]
Topic: Designing and building useful Legged Robots
CSIRO has undertaken legged robot research going on for ten years. Research has been broad, spanning the design and control of the mechanisms, as well as perception using legged robots. The talk will cover development of multiple legged robots and the learnings from lab experiments and field deployments.
BioRyan Steindl started working at CSIRO’s autonomous systems laboratory in early 2013 as a Mechatronics Engineer. He graduated Mechatronics Engineering at Queensland University of Technology (QUT) in 2013. His work at CSIRO heavily revolves around robotic system design with a focus on legged locomotion and robotics in manufacturing. Duties include operation of 3D printing laboratory for the Robotics and Autonomous Systems Group, and engineering support for research projects spanning multiple disciplines within CSIRO’s Data61.
Session-4:
15:30-15:35 Announcing the nominees for the best extended abstract award [21:30-21:35 CEST]
15:35-15:40 Oral presentation of Nominee 1 [21:35-21:40 CEST]
15:40-15:45 Oral presentation of Nominee 2 [21:40-21:45 CEST]
15:45-15:50 Oral presentation of Nominee 3 [21:45-21:50 CEST]
15:50-15:55 Announcing the winner of the best extended abstract award [21:50-21:55 CEST]
15:55-16:00 Concluding remarks [21:55-22:00 CEST]