I am an avid researcher specializing in **Computer Vision, Edge Computation, and Artificial Intelligence for aerospace applications**. I design guided flight control systems, GPS-denied autonomous drone navigation suites, and high-resolution satellite imagery analysis pipelines.
class AutonomyResearcher:
def __init__(self):
self.name = "Swarnim Bhadale"
self.focus = "Aerospace, Vision & GNC"
self.hw_targets = [
"Jetson Orin", "Pixhawk",
"Vicharak Axon", "Raspberry Pi"
]
def code_and_deliver(self):
return "⚡ Real-time Edge Autonomy"A showcase of advanced computer vision pipelines, guidance algorithms, and flight software architectures.
Developed Siamese ViT and GAN pipelines for multi-temporal satellite change detection on roads/buildings. Unified SuperPoint & LightGlue feature matching to align image pairs. Preserved spatial CRS metadata using Rasterio and TorchGeo, deploying ONNX Runtime quantization for memory-constrained, sliding-window inference.
Researched and built control systems for a small-scale model rocket to maintain heading using real-time Inertial sensors and GPS telemetry. Developed dead reckoning flight path logic by fusing INS and GPS systems for robust heading holds.
Developed custom perception and navigation algorithms enabling an autonomous UAV platform to dynamically avoid obstacles and identify mission-critical targets in real-time during execution.
Engineered state-machine flight controllers using real-time sensor feedback for unguided rockets, flight-validated to a peak altitude of 300m. Implemented 6-DoF trajectory estimators using MATLAB and Simulink.
Worked on NVIDIA Jetson, Raspberry Pi, Pixhawk (PX4), ROS, PyTorch, and ONNX to advance drone autonomy:
Published research articles, community outreach, and special awards.
Developed completely private and lightweight local cloud storage model based on the Vicharak Axon edge platform.
🏆 Best Paper Award (Security Track, Abhinav 2K26)Designed a predictive machine-learning neural model on educational data aggregates to flag at-risk dropout cases for early academic interventions.
Designed an active, real-time response pipeline utilizing anomaly classifiers to detect and mitigate DDoS attacks, improving overall network resilience.
Moments from rocket launch operations, field autonomy testing, and professional award distributions.
Photo
Swarnim (second from right) and the rocketry operations crew preparing experimental solid-motor payloads for launch.
Photo
Group portrait of the entire Team Anantam crew standing in the field next to the fully assembled, vertically mounted rocket on its launchpad just moments before launch.
Spectacular on-field flight footage capturing the countdown, high-thrust ignition, dramatic lift-off, and vertical ascent of the team's custom experimental rocket.
On-field footage of Swarnim verifying telemetry, ground station configurations, and flight software parameters on his laptop before ignition.
On-field footage of Team Anantam calibrating the launch guide rod and aligning flight telemetry trackers prior to ignition.
Raw footage of the rocketry crew conducting final system pre-checks and launchpad pre-flight configurations.
Photo
Swarnim (2nd from left) receiving the Best Paper Award at the Abhinav 2K26 young professional conference for his edge-computing private cloud backup design.
Photo
Close-up of your custom integrated flight computer bay, showcasing an ESP32 microcontroller, SD telemetry logger, solid-state sensors, and relays.
Data processing and control execution flow pipeline
I am actively seeking full-time roles and R&D opportunities in GNC Engineering, Flight Software, and Edge AI. If you're hiring, looking to collaborate on autonomous systems, or want to discuss aerospace innovation — let's connect!