BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:Physical Math Seminar
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20261015T173748Z
UID:tag:localist.com\,2008:EventInstance_44661600427031
DTSTART:20231031T183000Z
DTEND:20231031T193000Z
DESCRIPTION:Speaker:  Baxi Zhong  (Georgia Institute of Technology)\n\nTitl
 e:  Multi-legged matter transport:  A framework for locomotion on noisy la
 ndscapes\n\nAbstract:\n\nWhile the transport of matter by wheeled vehicles
  or legged robots can be guaranteed in engineered landscapes like roads or
  rails\, locomotion prediction in complex environments like collapsed buil
 dings or crop fields remains challenging. Inspired by principles of inform
 ation transmission which allow signals to be reliably transmitted over noi
 sy channels\, we develop a “matter transport" framework demonstrating th
 at non-inertial locomotion can be provably generated over “noisy" rugose
  landscapes (heterogeneities on the scale of locomotor dimensions). Experi
 ments confirm that sufficient spatial redundancy in the form of serially-c
 onnected legged robots leads to reliable transport on such terrain without
  requiring sensing and control.\n\n\nDespite robustness\, locomotors with 
 excessively redundant legs are often practically unfavored because of limi
 ted efficiency and applicability. Analogous to signal transmission\, we fu
 rther improve locomotion efficiency by properly coordinating (coding) the 
 redundant legs. The challenges of such coding partially lie on the high di
 mensionality associated with the additional legs and the emergent importan
 ce of inter-leg centralized coordination. Specifically\, we need a top-dow
 n approach to analyze the central coordination among the additional legs\,
  and further design how it should adapt to different environments. We use 
 geometric mechanics\, a mathematical framework for studying locomotion in 
 various systems\, for motion planning in multi-legged robots operating in 
 complex environments. As a result\, open-loop operation on multi-legged ro
 bots achieves remarkable performance on terrains with different types and 
 levels of complexity. Additionally\, analogies from communication theory c
 oupled to advances in coding for error detection/correction further improv
 e the locomotion efficiency and robustness via centralized adaptation (usi
 ng simple contact sensors to estimate environmental uncertainty). This res
 earch contributes to the field of legged robot locomotion\, providing new 
 possibilities for designing effective and adaptable robots for challenging
  environments.
GEO:42.358262;-71.090045
LOCATION:Building 2\, 449
SUMMARY:Physical Math Seminar
URL;VALUE=URI:https://calendar.mit.edu/event/physical_math_seminar_5635
CATEGORIES:Conferences/Seminars/Lectures
END:VEVENT
END:VCALENDAR