BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Cross Programme Talk: Explainable Augmented Intelligence (AI) for 
 Crack Characterization - Larissa Fradkin (Sound Mathematics Ltd)
DTSTART:20230512T100000Z
DTEND:20230512T110000Z
UID:TALK200380@talks.cam.ac.uk
DESCRIPTION:Crack characterization is one of the central tasks of NDT&E (t
 he Non-destructive Testing and Evaluation) of industrial components and st
 ructures. These days data necessary for carrying out this task are often c
 ollected using ultrasonic phased arrays. Many ultrasonic phased array insp
 ections are automated but interpretation of the data they produce is not. 
 This presentation describes an approach to designing an explainable AI (Au
 gmented Intelligence) to meet this challenge. It describes a C++ code call
 ed AutoNDE\, which comprises a signal-processing module based on a modifie
 d total focusing method that creates a sequence of two-dimensional images 
 of an evaluated specimen\; an image-processing module\, which filters and 
 enhances these images\; and an explainable AI module&mdash\;a decision tre
 e\, which selects images of possible cracks\, groups those of them that ap
 pear to represent the same crack and produces for each group a possible in
 spection report for perusal by a human inspector. AutoNDE has been trained
  on 16 datasets collected in a laboratory by imaging steel specimens with 
 large smooth planar notches\, both embedded and surface-breaking. It has b
 een tested on other similar datasets. The presentation discusses results o
 f this training and testing and describes in detail an approach to dealing
  with the main source of error in ultrasonic data&mdash\;undulations in th
 e specimens&rsquo\; surfaces.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR