The project aims to model the post-collision behaviour of collision-resilient quadrotors for collision-inclusive trajectory planning.• Design of experiments utilizing PX4 – ROS2 – Vicon Motion capture pipeline and deployed momentum-based external wrench estimation algorithm for collision detection on a custom-built collision-resilient quadrotor.• Achieved a 50% speedup in data collection by enhancing the onboard mission planning algorithm.

• Responsible for the design and development of a Quadcopter (Unmanned Aerial Vehicle). Roles included CAD modelling, rapid prototyping, sensor integration, testing, and troubleshooting.• Developed a cost-effective and lightweight prototype of a 3D printable quadrotor to be utilized to capture geo-tagged images using an onboard camera.

- Designed a Tri-rotor Unmanned Aerial Vehicle(UAV) using CAD modelling software SolidWorks.- Parameter optimization of the model-based high fidelity position and attitude controller with Matlab & Simulink.- Fabrication of the designed model using RAISE3D printer.- Design and development of the testing setup for tuning the UAV's controller parameters.

- Integrated a Deep Learning model (BlazePose) with Flask webapp to estimate the human pose in real time.- Deployed the webapp on Cloud computing platforms like Heroku, AWS and Google Cloud Platform.