Monday, August 30, 2021 - 11:30am to 12:30pm
Event Calendar Category
Uncategorized
Speaker Name
Samir Menon, Jonathan Kuck, Harry Zhe Su
Affiliation
Dexterity
Zoom meeting id
056129
Join Zoom meeting
https://mit.zoom.us/j/92328419690
Abstract
Transitioning robots from well-controlled lab environments to the real-world has been an outstanding challenge for decades. The challenges are numerous and span a variety of engineering disciplines in machine intelligence, simulation, modeling, algorithms, control, and robotic hardware & software. Dexterity is a robotics startup that has engineered and deployed robotic systems that can intelligently manipulate tens of thousands of items in production, reason about and operate in dynamic environments, collaborate with each other using the sense of touch, and safely operate in the presence of humans. Dexterity’s robots ship hundreds of thousands of units in packaged food and parcel warehouses each day and are in production 24/6. The data we collect has taught us what works in the real world and what doesn’t. The consequences of realizing even limited robot autonomy within semi-structured workflows are urgent and compelling---the logistics industry has a 300,000 job shortage and would be transformed if robots could fill the gap. This talk (by Samir, Jonathan & Harry) will discuss our technological advances to date, which are poised to deploy thousands of robots and present outstanding problems that need to be addressed before we can deploy millions of robots.
Biography
Samir is a founder and CEO of Dexterity, a company working on transforming robots from hard-wired automatons into intelligent collaborative helpers. He has also worked extensively on control theory, mathematical modeling, simulation, human motor control, and neural computation. He has designed and built numerous robots. Samir holds a master's and a Ph.D. in computer science from Stanford University, where he was a Stanford Interdisciplinary Graduate Fellow. Prior to that, Samir completed his Bachelors in Technology at the Indian Institute of Information Technology, Allahabad.
Jonathan works on optimization, machine learning, and perception at Dexterity. Prior to joining Dexterity, he completed his Ph.D. at Stanford, where he worked on machine learning and robotic perception problems bridging theory and application. His research has improved the speed/accuracy tradeoff of approximate probabilistic inference by orders of magnitude, merged the boundaries of classical inference techniques with deep learning, and bridged to core perception problems such as multi-object target tracking and object detection. He completed his bachelor's in engineering physics at UIUC.
Harry Zhe Su is a roboticist at Dexterity Inc. where he focuses on developing algorithms and robotic systems to enable robots to achieve dexterous manipulation. He is a senior architect for the robotics team and had a critical role in taking Dexterity’s intelligent dexterous robots into production in the parcel industry. Before joining Dexterity Inc., Harry received his Ph.D. in biomedical engineering from the University of Southern California (USC) by developing various biomimetic tactile perception algorithms and robotic manipulation algorithms using tactile sensing.