RocketNav — Rocket Trajectory Prediction Device
1240 1180 980 1040 1120 LAKE
▲ ROCKETNAV
T-0 · LAUNCH APOGEE
PREDICTED LZ
PROJECT STATUS: IN DEVELOPMENT

Predicting where a rocket lands — before it does.

RocketNav is a hardware + software system that predicts a model rocket's landing point in real time, during flight. An onboard flight computer streams motion, altitude, and position data to a ground station, where a machine-learning model plots the predicted landing zone on a map — with a live circle of uncertainty — while the rocket is still in the air.

OBJECTIVE → Know where the rocket will land before touchdown, so recovery is faster, safer, and never a guessing game.
SCROLL · PROJECT OVERVIEW ↓
WHY I'M BUILDING THIS

My story

{{ p.text }}
{{ storyPhotoEl }}
PHOTO · FROM SUPABASE

{{ storyCaption }}
PIPELINE

The process

01 · SENSE
Onboard sensor suite
MPU6050 IMU, BMP280 barometric altimeter, and GPS capture motion, altitude, and position from ignition to touchdown.
02 · FUSE
Kalman filtering
Raw sensor streams are fused into a clean, noise-reduced estimate of the rocket's true flight path.
03 · PREDICT
ML landing model
A model trained on real flight data forecasts the landing point and quantifies its confidence as an uncertainty radius.
04 · DOWNLINK
LoRa telemetry
A XIAO ESP32S3 with integrated LoRa radio transmits live to the ground station — predictions update mid-flight, not after recovery.
CREW MANIFEST

Team

{{ m.photoImg }}
PFP
{{ m.name }}
{{ m.role }}

{{ m.bio }}
FLIGHT HARDWARE
REV A
XIAO ESP32S3
Flight computer with integrated LoRa radio
MPU6050
6-axis accelerometer + gyroscope
BMP280
Barometric altimeter
GPS MODULE
Live position fix through descent
KALMAN FILTER
Sensor fusion for a clean state estimate
ML MODEL
Landing-point prediction, trained on real flights
FLIGHT LOG

Follow the build, entry by entry

Hardware revisions, software milestones, and flight tests are logged as the project moves forward.

VIEW THE TIMELINE →
▲ ROCKETNAV · ROCKET TRAJECTORY PREDICTION DEVICE PROJECT LOG · 2026