· Generate structured PLOT3D grids using Fidelity Pointwise with varying aft-end geometries for conic bodies· Run CFD simulations using LAURA to analyze the impact of aft-end geometry on hypersonic wake size· Evaluate CFD results to identify variations in wake temperature distributions, focusing on thermal signatures and ionization and recombination behaviors using Tecplot

· Simplified CAD geometry of the Dragonfly vehicle using Rhino3D and converted it into overset and multi-block grids using Chimera Grid Tools (CGT) and OVERGRID· Ran high-fidelity blade-resolved simulations using OVERFLOW for Dragonfly’s blade geometry and compared results with low-fidelity rotor-disk simulations using Tecplot· Evaluated different blade-tip meshes to find the best way to model blade-tip loss caused by blade-tip vortices· Analyzed forces and moments acting on Dragonfly’s fuselage from rotor-fuselage interactions using blade-resolved and rotor-disk methods, comparing results to that of a rotor interacting with a wall as the worst-case scenario· Currently investigating wind-tunnel wall effects on rotor aerodynamics to explain discrepancies between CFD and Dragonfly’s wind-tunnel experimental results

· Built and performed maintenance on the drones used for testing multi-agent localization and Simultaneous Localization and Mapping (SLAM) systems· Created a reinforcement learning guide on GitHub for onboarding: https://github.com/jace-kun/RL-Intro.git· Implemented a multi-agent tag game using Pytorch and OpenAI Gym in order to understand the multi-agent path planning problem

· Utilized recent advances in artificial intelligence and machine learning for optical navigation around the Moon· Leveraged Pytorch and MIT’s super computer, SuperCloud, to train a YOLO object detection algorithm to identify craters and other features on the surface of the Moon

· Assembled the avionics towers: flight computer, power distribution, camera board, telemega, batteries, etc.· Worked with Structures Sub-team to integrate the avionics towers with the rest of the rocket in preparation for launch in January 2023: https://youtu.be/qp7AeOoM4-4?si=NQ-OTh2gSC9QKmP3

· Generated Dragonfly rotor grids using Chimera Grid Tools (CGT) and OVERGRID· Ran CFD simulations using OVERFLOW CFD Flow Solver· Post-processed flow solutions using Tecplot360 and Python scripts

· Investigated ways to optimize the time needed to transfer satellite data to ground stations using a network of small optical relay satellites in low earth orbit· Utilized Cesium and Python libraries, Poliastro and Astropy, to create a user-friendly satellite constellation visualizer to help researchers understand satellite systems and their corresponding actions: propulsive maneuvers, remote sensing tasks, etc.

· Introduced to the theoretical foundations of machine learning· Leveraging KNN, logistic regression, decision tree, random forest, SVM, linear regression, and deep learning, created algorithms able to determine the release year of given songs based on data collected from music service Spotify using Pandas and Scikit-Learn· Compared these algorithms and presented results at MOSTEC Symposium, an event facing the broader MIT community