Building machines that sense, decide, and move
DHRUVDATTA
Mechanical +Manufacturing Engineer
Prev @NASA + GE Vernova
Hardware at NASA. Engines at GE.
- NASA Marshall Space Flight Center
- Built FEMAP finite element models for test hardware, verifying strength and structural margins against design requirements
- Analyzed welded and bolted joints using FEA reaction loads, FBDs, and hand calculations to validate mechanical hardware
- Supported mechanical hardware testing and validation alongside test engineers, comparing analysis with test performance
- GE Vernova
- Led design proposal addressing a 500 kW performance loss per engine, evaluating modifications to prevent $20M+ in losses
- Conducted CFD and flow simulations on 2 stages of GE aerospace engines to evaluate performance and design improvements
- Collaborated with design, systems, and performance teams to evaluate design feasibility and engineering tradeoffs
Every subsystem tells part of my story.
An interactive map of the engineering behind what I build.
- Hardware Design: experience with CAD, FEA, Mechanical Design, Joint Analysis. Skills: FEMAP, SolidWorks, Hand Calcs, CFD.
- Sensors: experience with Computer Vision, Cameras, Calibration. Skills: OpenCV, Python.
- Intelligence: experience with Learned Control Policies, Imitation Learning. Skills: PyTorch, CNNs.
- Compute: experience with Simulation, Compressible Flow, Modal Analysis. Skills: Python, MATLAB, Unity.
- Controls: experience with Closed-Loop Control, Robot Software. Skills: ROS 2, Python, Feedback Control.
Building Physical Intelligence
- Robotics Automation and Design Laboratory
- Built a ROS2 integrated physics simulation stack with URDF/XACROs to test robot behavior and system interactions
- Developed Drake based joint and task space control systems incorporating hydroelastic contact models, enabling successful sim-to-real deployment
- Developed coupled FEA/dynamics model for robotic satellite solar panels, improving deformation and response prediction
- Bio Nanorobotics Laboratory
- Designed and fabricated nanorobot prototypes in SolidWorks using UV resin additive manufacturing
- Developed magnetic actuation for drug delivery systems, optimizing nanoparticle alignment to improve actuator response
- Engineered a path-based control scheme for magnetically actuated nanorobots to optimize trajectory tracking
Physical AI: perception, policy, control and actuation in one closed loop
Selected Work
Selected work
Each project is its own world. Select one to travel there.
Contact
Let's build something that moves.
Open to internships, research and hardware-heavy problems.
hello@example.com