BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
 M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20260330T111912Z
DTSTART:20260415T100000Z
DTEND:20260415T110000Z
SUMMARY:AI-Fun & ELLIS Invited Speaker Series | Gabriella Pizzuto
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}k1vc-mn7khu
 pq-pbnjcw
DESCRIPTION:April’s AI-Fun and ELLIS invited speaker is Gabriella Pizzuto
  from the University of Liverpool. \n\nThese events provide an opportuni
 ty for anyone working in AI and AI-related research across the Universit
 y of Manchester to hear from top researchers about their current work. I
 n person attendance is encouraged. If you are unable to physically atten
 d\, please register via the Ticketsource link provided to receive the Te
 ams link to join online.\n\nTitle: Upskilling Robotic Chemists For AI-Dr
 iven Scientific Discovery\n\nAbstract: The demand for the rapid developm
 ent of new materials\, ranging from sustainable formulations to novel dr
 ugs\, requires a paradigm shift from manual experimentation to autonomou
 s discovery loops. While robotic chemists offer a path toward this goal\
 , their full potential remains untapped due to the inherent brittleness 
 of traditional automation in human-centric labs and the difficulty of mo
 delling complex robot-material interactions. In this talk\, I will discu
 ss our research on upskilling robotic systems through a new research are
 a we call Laboratory Skill Acquisition: the development of robust\, lear
 ning-based robot skills tailored for scientific laboratories. I will foc
 us on the challenges of performing complex\, contact-rich manipulation t
 asks with materials that push the boundaries of current simulation frame
 works\, such as granular powders and heterogeneous solids. Drawing on ou
 r recent work\, I will first demonstrate how embedding material-specific
 \, physical priors into the learning process can enable precise autonomo
 us material manipulation. I will then discuss our work in adaptive contr
 ol to handle the uncertainty of heterogeneous material properties when r
 etrieving samples from lab glassware. Furthermore\, I will briefly addre
 ss how failure-recovery mechanisms allow robotic scientists to adapt to 
 evolving workflows and operate safely alongside humans. I will conclude 
 by outlining the open challenges in this domain\, specifically focusing 
 on the underpinnings of safe\, human-in-the-loop robotic chemists and th
 e quest for robotic scientists capable of long-term autonomous discovery
 .\n\nBio: Dr. Gabriella Pizzuto is a Lecturer (Assistant Professor) in r
 obotics and chemistry automation at the University of Liverpool. Previou
 sly\, she worked as a senior postdoctoral research associate on the ERC 
 Synergy Grant 'Autonomous Discovery of Advanced Materials' at the Univer
 sity of Liverpool and as a postdoctoral research associate at the Univer
 sity of Edinburgh. She obtained her Ph.D. in computer science from the U
 niversity of Manchester in 2020\, where she was also a Marie-Sklodowska 
 Curie early stage researcher. Throughout her PhD\, she was a visiting sc
 holar at the Italian Institute of Technology and the Institute of Percep
 tion\, Action and Behaviour (University of Edinburgh). She is the recipi
 ent of a Royal Academy of Engineering Research Fellowship (2023) and EPS
 RC New Investigator Award (2025) towards advancing robotic chemists for 
 acquiring new laboratory skills. She is a research area lead at the UK's
  National Institute for Advanced Materials R&I (Henry Royce Institute) a
 nd co-chair of the ECR committee of the UK’s AI in chemistry hub. Her pi
 oneering work has been awarded an Outstanding Paper in Automation finali
 st at IEEE ICRA 2022 and a Best Healthcare Automation Paper finalist at 
 IEEE CASE 2024. Her current research interests are contact-based and phy
 sics-constrained robot skill learning\, failure recovery methods in labo
 ratory environments and safe human(chemist)-robot collaboration. \n
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:Lecture Theatre 1.4\, Kilburn Building\, Manchester
END:VEVENT
END:VCALENDAR