02 / Projects

Things I have made, and shipped.

Selected engineering and research projects in autonomy, vehicle control, network design, and race strategy.

05/2025 - ONGOING

Model Evaluation for Autonomous Vehicle Systems

Goal: Evaluate vehicle-dynamics models using real and simulated data for safety, performance, and scalability in system-of-systems autonomous driving.

Technical work: Applied functional analysis methods to model evaluation and integrated performance metrics into system-of-systems tradespace evaluation for multi-agent strategy optimization.

Skills: MATLAB, Simulink, vehicle dynamics, functional analysis, model evaluation, system-of-systems analysis, multi-agent optimization.

Major achievement: Built an evaluation approach connecting model performance metrics to autonomous system tradespace decisions.

08/2023 - ONGOING

Uncertainty Mitigation for High-Speed Off-Road Autonomy

Goal: Improve high-speed off-road autonomy when model errors and unmatched uncertainties affect vehicle behavior.

Technical work: Developed an adaptive control method to operate alongside a nominal model predictive controller and compensate for model mismatch.

Skills: Adaptive control, model predictive control, robust control, optimal control, vehicle dynamics, MATLAB, Simulink.

Major achievement: Created a combined nominal-MPC and adaptive-control architecture for uncertain off-road vehicle operation.

02/2023 - 05/2023

Path Planning in the Presence of Mobile Threats

Goal: Navigate an ego aircraft through environments containing static and stochastic mobile threats.

Technical work: Developed a dynamic-programming motion-planning control law and treated collisions or failure to reach the goal as unsuccessful outcomes.

Skills: Dynamic programming, motion planning, control-law design, probability and statistics.

Major achievement: The planner achieved a 57% success rate under the project evaluation criteria.

01/2023 - 03/2023

Optimal Race Strategy Estimation Algorithm

Goal: Determine optimal pit-stop timing and lap strategy under stochastic tire degradation.

Technical work: Built a dynamic-programming algorithm that penalizes pit time, models uncertain tire wear, and rewards remaining on track with well-preserved tires.

Skills: Dynamic programming, probability and statistics, stochastic modeling, optimization, MATLAB.

Major achievement: Produced an algorithmic race-strategy estimator that balances pit time against uncertain tire performance.

02/2022 - 05/2023

Distributed Design of Controllable and Robust Multi-Agent Networks

Goal: Design multi-agent networks that are both completely controllable and maximally robust with minimal inputs.

Technical work: Developed a decentralized network-design algorithm that optimizes the opposing properties of controllability and robustness.

Skills: Strong structural controllability, zero forcing, graph grammars, graph theory, robust network design, multi-agent systems, algorithm design.

Major achievement: Obtained a polynomial-time optimal solution to a problem that is otherwise NP-hard.

01/2020 - 07/2020

Compliant Mechanism for Wing Actuation

Goal: Reduce the design and manufacturing complexity of a vehicle wing-actuation mechanism.

Technical work: Engineered a four-bar compliant mechanism using kinematic synthesis, then performed load testing and simulation.

Skills: Kinematic synthesis, CATIA, ANSYS, mechanical design, structural simulation, load testing.

Major achievement: Reduced part count while validating the mechanism through ANSYS simulations and load testing.

08/2017 - 06/2018

Control System Synthesis for Autonomous Mobile Robot - ADAM

Goal: Enable accurate torque vectoring, velocity control, and trajectory tracking for a differential-drive autonomous robot.

Technical work: Modeled longitudinal and lateral dynamics and designed a PID controller in MATLAB and Simulink.

Skills: PID control, vehicle modeling, longitudinal and lateral dynamics, torque vectoring, trajectory tracking, MATLAB, Simulink.

Major achievement: ADAM placed 9th in IGVC 2018; its successor SOLO won IGVC 2019 at Oakland University.

03/2017 - 02/2019

Vehicle State Prediction Model of Autonomous Mahindra e2o - EVE

Goal: Develop the vehicle model, localization, and simulation foundation for an autonomous Mahindra e2o.

Technical work: Localized vehicle position from encoder readings, used the state estimate for control, designed the vehicle in CATIA with surface modeling, and fabricated and assembled the conversion.

Skills: Vehicle state estimation, encoder-based localization, CATIA, surface modeling, vehicle simulation, fabrication, assembly.

Major achievement: Contributed to building India's first SAE Level 3 autonomous vehicle through Project Manas; the CATIA model became the primary simulation model.