AGL operates Australia’s largest private electricity generation portfolio within the National Electricity Market, comprising coal and gas-fired generation, renewable energy sources such as wind, hydro and solar, batteries and other firming technology, and gas production and storage assets.
AGL’s goal to consistently innovate in energy to enhance the way Australians live and to help preserve the world around us for future generations, was a perfect fit for APR’s to support the growth of a diverse workforce and accelerate innovation.
APR Intern Mohsen Ghofrani-Jahromi, from Monash University, was provided with the opportunity to work on training and deploy deep learning models for the automatic analysis of drone imagery at AGL.
As a PhD student at the Turner Institute for Brain and Mental Health, Mohsen’s research focussed on using machine/deep learning to enhance prognosis and stratification in Huntington’s disease. These models would ultimately improve how clinical trials are being conducted.
“During my internship, my main focus was on training and deploying deep learning models for the automatic analysis of drone imagery. The primary objective was to have a computer vision tool that could identify objects of interest in AGL’s power stations.”
– Mohsen Ghofrani-Jahromi, Former APR Intern at AGL.
Image L-R: Matthew Lee Chief Remote Pilot and Drone and Robotics Engineer, AGL. Rahul Rao, Analytics Engineer, AGL. Mohsen Ghofrani-Jahromi, APR.Intern. Supplied.
Mohsen credits his AGL mentors who guided him to train a prototype object detection model.
“Being able to provide input to AGL’s chief drone pilot regarding the requirements for images to improve the training of models, was one of my biggest achievements,” Mohsen said.
“Because, AI does not work well, unless you feed it with sufficiently good data, free of noise as much as possible.”
“APR.Intern connected us with PhD candidates and their academic mentors, allowing us to gain access to state-of-the-art knowledge in machine learning techniques for computer vision. Mohsen’s contribution has greatly expanded the expertise of our team in qualifying use-cases for object detection, developing a system for labelling images and training models, and evaluating the efficacy of trained models.”
– Rahul Rao, Analytics Engineer, AGL.
Image L-R: David Bowly, Manager, Engineering Analytics, AGL. Rahul Rao, Analytics Engineer, AGL. Amir Saberi, Analytics Engineer, AGL. Mohsen Ghofrani-Jahromi, APR.Intern. Nellie Georgiou-Karistianis, Pro Vice-Chancellor (Research Training), Monash University. Susmita Saha, Research Fellow, Turner Institute for Brain and Mental Health, Monash University. Supplied.
This four-month internship to join AGL gave Mohsen confidence to excel in industry for years to come. “I hope to have explored various areas of AI in ten years’ time and I am pretty sure this internship was a decent first step.”
