Undergraduate Student at Indian Institute of Technology – Jodhpur | Mechanical Engineering | Data Driven Fluid Mechanics | Passionate Wildlife Photographer.
About my work and research …
My biggest interests lie in Computational Mechanical Engineering, especially in the field of combining fluid mechanics and machine learning. The use of biomimetic design for aerodynamics has been my forefront research field. I have been working on parametric space exploration using machine learning for design optimization problems in the field of fluid mechanics. Data driven analysis of fluid systems has also been a major working interest of mine. My recent research work involves venturing into the domains of physics based machine learning.
Machine Learning
Aerodynamics
Biomimicry
Academic Record …
My cumulative grade point average after the 7th semester is 9.52/10.
My Department Rank is 1 (1st out of 64 students in Department of Mechanical Engineering) and the Institute Rank is 1 (1st out of 450 students in among all departments).
Publications
Rushil Samir Patel, & Harshal D. Akolekar, (2022). Machine-Learning Based Optimisation of a Biomimiced Herringbone Microstructure for Superior Aerodynamic Performance. bioRxiv. doi:10.1101/2022.09.17.508361
Rushil Samir Patel, & Harshal D. Akolekar, (2022). Supervised Learning Augmented Computational Fluid Dynamics for Bio-inspired Herringbone Structure Optimisation. (Paper accepted in International Conference on Recent Advances in Mechanical Engineering, 2022) (Link)
Awards
Best Paper of the Day Award.
For the paper titled “Supervised Learning Augmented Computational Fluid Dynamics for Bio-inspired Herringbone Structure Optimisation” in International Conference on Recent Advances in Mechanical Engineering, 2022.
Frederic Barnes Waldron – Best Student Award (2022)
The Frederic Barnes Waldron ‘Best Student’ Award is awarded each academic year on the recommendation of the Head of Department responsible for undergraduate degree courses (here Mechanical Engineering) that are accredited by the Institution. It is awarded for overall meritorious performance not only in academics but also for merit in other activities concerning the said engineering discipline including (but not limited to) research and projects.
Research highlights
Machine Learning Augmented Optimisation of biomimiced structures For Superior aerodynamic Performance
Working under the guidance of Dr. Harshal D. Akolekar (Indian Institute of Technology, Jodhpur; University of Melbourne, Australia) to create a novel semi-automated workflow to optimise a biomimiced herringbone riblet microstructure for better aerodynamic performance and drag reduction using machine learning algorithms.
Digital Identification Of Interacting Shockwave Stems Using Machine Learning Algorithms
Worked under Dr. Li Xi at McMaster University, Hamilton, Canada. The work is based on using machine learning to understand turbulent flow and the physics behind vortex dynamics. The main aim is to work on high fidelity DNS data for wall bounded flows using deep learning and artificial intelligence.
Turbulence Recognition using deep learning tools for flow over rough plates.
Collaborating with Dr. V N Rao (Indian Institute of Technology, Madras); under Dr. Harshal D. Akolekar (IIT Jodhpur; University of Melbourne, Australia) on CNN autoencoder based low – dimensionalization methodology for fluid flow data on rough plates collected at Dr. V N Rao’s Lab. We also aim to identify the exact location of the onset of turbulence within the transition region using DL methods.
Digital Identification Of Interacting Shockwave Stems Using Machine Learning Algorithms
Working to identify the shockwave stems in normal shocks using machine learning approaches from Schlieren images of intermixing supersonic flows.
Deep learning and graph network based paradigm for multi-objective control of unsteady fluid flows.
Working with Dr. Aditya G. Nair (University of Nevada, Reno – United States of America) on creating autoencoder based flow decomposition systems for understanding interactions between flow states within a latent space upon initial perturbations in flow.
Projects
Shockwave detection using project orthogonal decomposition and machine Learning.
- Obtained Schlieren images of shock waves reflected from double wedge system. Reduced dimensions of the obtained images for usability.
- Found different modes of shockwave structure by applying proper orthogonal decomposition by employing the covariance matrix, SVD and reduced SVD methods.
- Used pre-trained SRGAN model for super-resolving the derived modes from a compressed image into a high resolution image.
- Applied the aforementioned K-Means clustering based saliency map algorithm from machine learning to separate the shock structures from background.
- Used Canny Edge Detection algorithm to generate isolated shock structure. Applied Hough Transformation to find the shockwave angle to find the post shock mach number without physical experimentation.
GUIDELINE DRIVEN PHOTOGRAPHY BY ROBOTS USING MACHINE LEARNING.
- Presented a critical review (Link), extended abstract (Link) and technical update (Link) on an existing research paper of photography by robots.
- Tested and coded a new and efficient saliency map generation method using K – Means Clustering – an unsupervised learning algorithm in machine learning.
- Proposed a method for separating noise from an image using the generated saliency map before image classification.
- Ideated a faster method of classification based on using different machine learning models depending on different data parameters.
E-BAJA SAE INDIA NATIONAL COMPETITION
- Responsible for designing the main Roll Cage of an all wheel drive vehicle. Used AutoCAD and SolidWorks for this process. Iterated over several self made vehicle designs based on paper readings and competition rule book to find design loop holes.
- Completed the static analysis of the selected optimised Roll Cage in Ansys Static Structural. Verified the safety of the said Roll Cage.
- Represented Roll Cage and Ergonomics subsystem in virtual rounds. Was responsible for presenting a detailed review of the Roll Cage subsystem to a panel of judges.
- Manufactured the Roll Cage of an all terrain vehicle. Learnt operating manufacturing equipment such as lathe, cutting tool, grinder etc.
Skills
| PROGRAMMING | MACHINE LEARNING TOOLS | SOFTWARES | OTHERS |
| Python (Jupyter and Colab) | Numpy | ANSYS Fluent | Adobe Lightroom |
| C | Scikit Learn | ANSYS Static Structural | Adobe Photoshop |
| C++ | Pandas | AutoCAD | Adobe Illustrator |
| LaTeX | Pytorch | SolidWorks | |
| MATLAB * | Creo* |
Responsibilities
Joint Secretary
of Shutterbugs, The Photography and Photo-Editing Society, IIT Jodhpur (de facto Board Member of the Board of Art and Culture (BAC), IIT Jodhpur).
Relevant Links
- Mail : patel.28@iitj.ac.in
- Google Scholar : scholar/rushilpatel
- ORCiD : 0000-0003-4416-7577
- LinkedIn : linkedin/rushilpatel
- Instagram : instagram/rushilpatel
Rushil Samir Patel 