Shanti Swaroop K
- Research Program Mentor
PhD at Indian Institute of Technology, Hyderabad
Expertise
AI/ML, Mechanical and Aerospace Engineering, Data Science, Renewable Energy, Electrical Engineering, Astrophysics
Bio
My main interest is in understanding complex systems—from the human body to rockets—using mathematical equations and physics. I’m especially fascinated by modeling, where I combine data-driven approaches with physics-informed machine learning to uncover new solutions to challenging problems. Whether it’s how the brain maintains balance or how rockets escape Earth’s gravity, I love digging into the math and dynamics behind it. Outside of my research, I really enjoy working with students. They ask the toughest questions and keep me on my toes, constantly pushing me to learn and think in new ways. It’s a rewarding part of my life that keeps the spark of curiosity alive.Project ideas
Modeling the Brain as a Balance Controller: Insights into Traumatic Brain Injury
The human brain acts as the control center of the body, constantly processing sensory input and issuing commands to maintain balance, coordination, and movement. When we walk, stand, or recover from a fall, the brain integrates information from the inner ear, eyes, and muscles to keep us upright. But what happens when this delicate system is disrupted—such as in a traumatic brain injury (TBI)? In this project, students could model how the brain functions as a control system in human balance and coordination. They could simulate or analyze how the brain responds to destabilizing inputs and what happens when parts of this system fail. This could lead to exploring how certain injuries affect balance, how protective equipment (like helmets) might reduce TBI risk, or even how machine learning can help detect early signs of instability. The project could culminate in modeling recovery trajectories, identifying high-risk scenarios, and suggesting real-world applications in sports, vehicle safety, or elder care.
Delayed Outbreaks — Modeling Disease Spread with Time Lags
Infectious diseases often don’t cause symptoms or transmission immediately after exposure. There’s a built-in delay between when a person gets infected and when they become contagious—a feature traditional models often overlook. In this project, students can explore how these time lags affect the spread of diseases by incorporating delay differential equations into classical epidemiological models like the SIR framework. By simulating diseases with different incubation periods, students can investigate how delays influence outbreak size, timing, and the effectiveness of interventions. This project offers a hands-on opportunity to connect real-world public health issues with mathematical modeling, with possible extensions into policy planning, vaccination strategy, or even real data fitting from recent pandemics.
Thinking on Your Feet — Modeling Balance with Neural Delays
The human body is a marvel of real-time control, especially when it comes to maintaining balance. But unlike machines, the brain doesn’t respond instantly—there’s always a brief delay between sensing instability and correcting it. In this project, students will model human balance using an inverted pendulum system with delayed feedback, mimicking the neural processing times in the brain and spinal cord. Using delay differential equations, they can simulate how small changes in reaction time or control strength can lead to tipping, wobbling, or stability. This project not only ties together physics, biology, and math but also opens the door to applications in injury prevention, robotics, physical therapy, or even the development of fall detection systems.
Traffic Jams and Human Delay — Modeling Stop-and-Go Waves on Highways
Ever wonder why traffic jams happen even without accidents or lane closures? One key reason is the reaction delay of drivers—how long it takes a person to respond to the car in front slowing down. In this project, students can model traffic flow using a chain of cars, each governed by a simple control rule with a built-in delay to mimic driver reaction time. Using delay differential equations, students can explore how different delay lengths and driving behaviors (like aggressive or cautious following) influence the formation of stop-and-go waves. The project could be extended to explore the impact of autonomous vehicles with zero or reduced reaction times, and how they might stabilize traffic flow in the future.
Delayed Climate Feedback — Modeling Earth's Temperature Control Loop
Earth's climate system functions like a giant control loop, trying to maintain energy balance. But feedback mechanisms—like the melting of ice reducing reflectivity or increased CO₂ trapping more heat—don’t take effect immediately. In this project, students can model climate regulation with feedback delays using delay differential equations to capture how temperature responds to CO₂ increases with a time lag. This simplified model can demonstrate how even small delays in feedback loops can lead to instability, overshoot, or runaway warming. It’s a great opportunity to apply control system ideas to environmental science and helps students see the value of systems thinking in solving global problems.