Women in Big Data Global

Donate
×

Blogs

DataCamp Donates & WiBD : AI for Utilities

Women in Big Data

By Modupe J. Owolabi,

April 30, 2024

image for 4-26 meetup

The Women in Big Data (WiBD) and DataCamp Donates monthly Zoom Info-Session took place last Friday. It was a very rich and informative talk given by Dr. Sridevi Pudipeddi. She explained the highlights of a Deep Learning project on which she has been working recently. It was really neat.

Introduction

The zoom meeting started with a warm welcome by Srabasti Banerjee, and a brief introduction to the world of Women in Big Data by Shala Arshi. Shala gave an overview of the WiBD website and the many fascinating programs going on. For newcomers this was sure to be inspiring, and for older hands it was a good reminder of how WiBD has helped many of us through the mentoring program, the leadership, trainings, and more recently the hackathons and podcasts. She then focused on the wonderful partnership WiBD has with DataCamp. Through this collaboration members have an incredible opportunity to learn and apply for certifications by going through DataCamp courses covering specific skills and career tracks. Members were encouraged to take advantage of the wide array of courses, and specialized training as well as the Associate and Professional Certifications, in Data Analysis, Data Science, and Data Engineering. Special recognition was given to Khadija Z who has now become the very first WiBD member to earn a certification on DataCamp. Khadija completed 86 hours of e-learning, covering 22 courses and 11 projects all while making top scores in the skill assessments. Indeed, she has made all WiBD DataCamp learners proud and has really set the bar high! Congratulations to Khadijah! Srabasti then introduced the top ten learners for the month and encouraged the learners to endeavor to write out their experience and post on LinkedIn or just send a direct email. This would serve to encourage and motivate others.

Our Guest Speaker

Next Srabasti introduced the guest speaker for the session Dr. Sridevi Pudipeddi. Sridevi holds a Ph.D. in mathematics and is an associate professor at Kansas State University teaching data science, machine learning and autonomous systems. Fun fact: Sridevi has a cute dog who is well looked after by Srabasti’s two lovely daughters! Also, Srabasti had been Sridevi’s student at some point. Kansas State University has an interesting campus at Salina. The Salina campus covers Aerospace management, carries out pilot training, has a 1-mile airstrip and promotes partnerships with some major industries. Sridevi’s courses are actually online and include machine learning and autonomous systems which are available in Bachelor’s, Masters, and Certificate degrees. She explained that not many universities in the U.S. are actually offering autonomous systems, but there are industries implementing autonomous systems like John Deere – for tractors, and others for fruit-picking machines and others for weeding.

AI for Utilities

Then Dr. Sridevi described the collaborative work on the project which covered Data Acquisition, Data preparation, Data reception, and Computational challenges. She explained how this was a capstone project that students would need to do a presentation on. She shared the need utility companies have to carry out regular maintenance, and woodpecker damage is a major problem that results in energy loss. (Interestingly, female woodpeckers are the major culprit.) The project involved using drones with cameras attached to get images of woodpecker damages to utility poles. Data teams would carefully follow through the data acquisition, noting the importance of data labelling and how much activity goes into the data preparation. 

    

Quite a lot of technical detail was explained in a simple style including the fact that many layers are needed to carry out a deep learning project. For this project the architecture used needed to process images, hence Convoluted Neural Networks (CNN). The model used was VGG16 which had 16 layers to process the data from the input layer to the output layer. It was clear that the amount of work involved based on the variations in images, the size of images, the actual sub-images needed (whether or not, and precisely where there is a hole of a specific diameter on the pole) was very detailed.  All this culminated in a large dataset. On the whole, it was very enlightening.  

Reactions and Conclusion

Certainly, this was an intensive project and we were all quite pleased, I’m sure, to be able to grasp the concepts and challenges involved. It was a real eye-opener, as to one of the practical ways AI is being used, and the need for more projects like this that affect communities directly. Definitely an enlightening session, and inspiring too. Very glad that Dr. Sridevi Pudipeddi is on our side! Hoping to see more Women in Big Data projects as we continue learning from the DataCamp courses.

Link to the recording.