Yash Toshniwal

Guiding Professor and University

Suhas Jain, Georgia Tech

Project Title

ML-Based Solution for Pressure Poisson Equation

Field

Fluid Mechanics and ML

How did you reach out to the prof?

I reached out to the professor via email. Since I had recently completed a course in fluid mechanics, I found that his research interests aligned closely with mine, which motivated me to contact him.

How did you prepare for your first meeting with the prof? What previous knowledge was required?

The professor was working on applying machine learning to fluid mechanics. To prepare, I went through his published research and lab work, and refreshed my understanding of core fluid mechanics concepts. I also brushed up on basic machine learning theory. The meeting itself was very relaxed, he was mainly looking for enthusiasm, along with a fundamental grasp of fluid mechanics and ML.

What was the research topic? How did you arrive at it? (did the prof give it/did you explore & find it yourself)

The chosen research topic was “ML-Based Solution for the Pressure Poisson Equation.” The professor had the initial idea and suggested a related research paper. After reviewing it, I found the topic engaging, so in our next meeting we agreed to work on this direction.

What was the duration of the project?

The project lasted for about two months, spanning the summer.

What was the work done by you as part of the project? Who were the other stakeholders involved?

Initially, I started working alone, but later invited my classmate Kosmika to collaborate, as handling the workload solo proved challenging. Together, we developed datasets and built a preliminary algorithm incorporating PCA and LSTMs into the solver’s framework. However, due to the remote nature of the project and the semester resuming, we couldn’t fully implement the solution.

What was the mentorship style like and how often did you meet with the professor?

The professor maintained a very relaxed mentorship style. We met once a week, but the project was largely self-paced. He provided guidance whenever we had doubts, giving us the freedom to explore and progress independently.

How was the overall experience? (Key learnings & Challenges)

It was a great experience. I got the opportunity to explore cutting-edge research at the intersection of mechanical engineering and machine learning. The biggest challenge was the remote setup, which limited progress compared to what might have been possible in an offline environment. Nevertheless, it was highly rewarding and aligned well with my interests.

What are the future plans following this project?

There were no concrete follow-up plans after this project.

What would you tell someone who is considering their first research project?

Choose a field you’re genuinely interested in, because research often requires deep reading and theoretical understanding. Embrace the process, even the setbacks, since not every project leads to perfect results. Enthusiasm, curiosity, and persistence go a long way in making the experience fulfilling.