BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Department of Materials Science & Metallurgy Goldsmiths' Seminars:
  AI\, microscopes\, and the quest for better materials - Prof. Sergei V. K
 alinin\, Department of Materials Science & Engineering\, University of Ten
 nessee
DTSTART:20250515T140000Z
DTEND:20250515T153000Z
UID:TALK232114@talks.cam.ac.uk
CONTACT:Caroline Teagle
DESCRIPTION:This talk will be delivered via Zoom\, but we will convene in 
 Goldsmiths' Lecture Theatre 2 where tea and cake will be provided\, so ple
 ase join us there!\n\nIn this presentation\, I will illustrate how ML-driv
 en electron and scanning probe microscopies can be leveraged to uncover st
 ructure-property relationships in complex materials\, extract fundamental 
 physical laws governing ferroelectric polarization dynamics and property e
 volution across combinatorial libraries\, and even manipulate matter on th
 e nanometer and atomic level. Central to this approach is the concept of p
 robabilistic reward functions\, which enable autonomous research workflows
  while integrating human-in-the-loop decision-making. I will demonstrate h
 ow reward-based automated characterization can be used to close the materi
 als discovery loop\, orchestrate diverse characterization tools across sha
 red chemical spaces\, and co-navigate costly experiments and epistemic unc
 ertainty-aware theoretical models. The special case here is operationalize
 d materials and physics discovery in combinatorial libraries\, where ML-en
 abled scanning probe microscope autonomously performs topography and spect
 roscopy tuning and combinatorial space exploration. I further discuss stra
 tegies to extend these strategies towards electron microscopy bypassing th
 e sample preparation bottleneck. Looking ahead\, this work lays the founda
 tion for the automated lab of the future\, where human intuition and AI-dr
 iven autonomy work in synergy to drive materials discovery at an unprecede
 nted scale.
LOCATION:Goldsmiths' 2 & https://zoom.us/j/96836714124
END:VEVENT
END:VCALENDAR
