WORKSHOP

Advancements in Trajectory Optimization and Model Predictive Control for Legged Systems

2nd Edition

event
17th May 2024

ABOUT

Robotic technology has proven to be an excellent solution for aiding humans in an ever-increasing number of scenarios: from domestic and urban routines to industrial tasks, robots reduce the workload burden on humans and their exposure to hazards. Despite the capabilities demonstrated in recent years, many obstacles remain to be overcome. New challenges arise from the increasing capabilities of advanced robotic platforms. The more complex and unstructured the environment, the more robots should be versatile and reliable to overcome obstacles, plan optimal motions, and reliably accomplish the designed tasks.

pic1

This workshop continues to explore state-of-the-art advancements in control strategies that are behind the abilities of such robots, namely, planning and control of dynamic, whole-body motions. Pursuing the thread initiated with the first edition of this workshop, we will focus on trajectory optimization and optimal control: successful approaches that exploit the robotic systems' dynamics, particularly under-actuated ones. This second edition will also address links and synergies with machine learning approaches, e.g. reinforcement learning or deep learning, which are undeniably increasing their relevance and popularity for planning and control applications. Both strategies boil down to a minimization (or maximization) of a desired metric, resulting in a sequence of the most effective actions to take. However, they are fundamentally different approaches that exhibit inherent advantages and drawbacks. How do robotic systems benefit from these methods? Which are the properties shared by the two? How to combine these strategies? These questions will be addressed, soliciting a discussion to compare ideas and solutions from the invited speakers.

pic1
pic2

PROGRAM

Time Section Speaker
8:45 - 9:00 Welcome Organizers
9:00 - 9:30 Talk 1 Andrea Del Prete
9:30 - 10:00 Talk 2 Carlos Mastalli
10:00 - 10:30 Talk 3 Johannes Englsberger
10:30 - 11:00 Coffee Break
11:00 - 11:30 Talk 4 Wang XingXing
11:30 - 12:00 Talk 5 Patrick Wensing
12:00 - 13:00 Poster Session Accepted Authors
13:00 - 14:00 Lunch Break
14:00 - 14:30 Talk 6 Jean-Pierre Sleiman
14:30 - 15:00 Talk 7 Yuval Tassa
15:00 - 15:30 Talk 8 Serena Ivaldi
15:30 - 16:00 Coffee Break
16:00 - 16:30 Talk 9 Pierre-Alexandre Léziart
16:30 - 17:15 Final Discussion Organizers + Speakers
17:15 - 17:30 Best Poster Award
20:00 - 22:00 Social Dinner Organizers + Speakers + Finalists

SPEAKERS

Click the play button on the top-right of each card to see the recorded talk!
Andrea Del Prete Picture

Andrea Del Prete

Università di Trento

Title: Integrating learning and trajectory optimization to achieve safe and efficient robot control

Carlos Mastalli Picture

Carlos Mastalli

Heriot-Watt University

Title: Unlocking Agile and Object-Aware Loco-Manipulation

Johannes Englsberger Picture

Johannes Englsberger

DLR

Title: Why I decided NOT to use MPC – a denier’s perspective

Wang Xingxing Picture

Wang Xingxing

Founder & CEO of Unitree

Title: AI-powered humanoid robot

Patrick Wensing Picture

Patrick Wensing

University of Notre Dame

Title: Accelerating MPC for dynamic locomotion: Exploiting structure, and letting learning guide the way

Jean Pierre Sleiman Picture

Jean Pierre Sleiman

ETH

Title: Whole-Body Motion Planning and Control for Multi-Contact Locomanipulation

Serena Ivaldi Picture

Serena Ivaldi

Inria

Title: Teleoperating humanoid robots: optimised controllers, contacts and human-like motions

Yuval Tassa Picture

Yuval Tassa

Google DeepMind

Title: Trajectory optimization with MuJoCo

Pierre-Alexandre Leziart Picture

Pierre-Alexandre Léziart

JRL Tsukuba

Title: Artificial locomotion: should we learn or optimize?

ORGANIZERS

SPONSORS