CARTE Industry Speaker Seminar Series welcome Jonathan Armitage, Lead Machine Learning Developer at AltaML, for an in-person seminar.

Topic: Visual Crack Detection via Rectangular Stitching

Date and Time: Friday March 3, 2023 (12:00 – 1:00 PM EST)

Speaker: Jonathan Armitage, Lead Machine Learning Developer at AltaML

Moderator: Professor Daniela Galatro, University of Toronto

Registration: Please register by filling out this form. Capacity is limited. Please register early to avoid disappointment.

Venue:  Sandford Fleming Building (10 King’s College Rd, Toronto, ON M5S 3G4), Room 3201

Abstract: Safety is a top priority in the oil & gas industry. This discussion will outline a product that was built through joint collaboration between AltaML and Kleinfelder and that identifies areas within a refinery where the fireproofing could become hazardous. We will walk through the process of going from “that is an interesting idea” to a deployable solution. We will detail the dataset (two dimensional spherical images), the applied computer vision technology, and how AltaMLs methodology played an integral role in balancing business and technical requirements while developing this solution. We will also discuss challenges in operationalization of this use case and how our MLOps team tackled it.

Speaker Bio: Jonathan Armitage is the Lead Machine Learning Developer at AltaML. Jonathan has over 10 years of applying statistical and machine learning in industry. He brings first-hand experience of building and deploying ML solutions within the FinTech, Healthcare and Sports industry. Jonathan holds a MSc in Economics from Florida Atlantic University and in a previous life Jonathan played professional baseball in the San Francisco Giants organization. The perspective and outlook that he applies when taking on a ML project have been curated by the lessons learned in the highly competitive environment of professional sports as well as those learned working in Silicon Valley start-ups. He is very passionate about applied ML as well as explainable AI, cost-sensitive ML and user adoption.