How It Works
The Science Inside Your Wing
Every vario measures the air around you. ParaBaro is the first instrument that measures the forces inside your canopy.
The Problem
Every existing paragliding instrument measures external conditions — altitude, GPS position, vertical speed, wind direction. None measures what happens inside the wing itself. Yet collapses — the primary safety hazard in paragliding — are caused by localised internal pressure loss.
For decades, pilots have relied on feel, experience, and intuition to assess their wing's condition. While these skills are invaluable, they're subjective, difficult to teach, and provide no record for post-flight analysis. ParaBaro was designed to close that gap — making the invisible visible.
The Physics
Your wing generates lift through differential pressure — higher pressure on the bottom surface, lower pressure on top. This pressure difference is what keeps you in the air, and monitoring it reveals critical information about wing loading, inflation state, and collapse risk.
Unlike absolute pressure (which primarily indicates altitude), differential pressure shows you the actual forces acting on your wing. A healthy wing maintains consistent differential pressure during flight. When this pressure drops — especially asymmetrically — it indicates reduced lift and potential collapse conditions.
This 45 Pa difference is generating lift. If it drops to 20 Pa, your wing is approaching collapse conditions. ParaBaro monitors this continuously across multiple wing positions.
Wing collapse — the event ParaBaro is designed to predict and prevent
Sensor System
ParaBaro measures the pressure difference between the interior of the wing cells and the ambient air using paired differential pressure sensors at multiple positions across the span.
All sensor channels are time-synchronised with competition-grade IGC GPS logging, creating a unified dataset of position, motion, and internal wing state — approximately 720,000 sensor readings per hour of flight.
Findings
Pressure asymmetry between left and right sensors diverges 0.5–1.5 seconds before a visible collapse. The signal is consistent and measurable above turbulence noise.
Thermal entries create a characteristic pressure gradient across the wingspan. Because the wing is wider than a thermal's shear zone, one side feels the thermal 1–2 seconds before the other.
By tracking baseline pressure signatures across flights, we can detect gradual changes in canopy behaviour — potentially identifying material fatigue or porosity changes before they affect safety.
Wing Classifications
Different wing certifications exhibit different pressure characteristics. Higher-performance wings (EN-C/D) operate with lower margins and require more active piloting.
Beginner/Leisure
Typical Range: 60-100 Pa
Cruise: 75-85 Pa
Collapse Threshold: ~35 Pa
High inherent stability with large safety margins. Wing naturally maintains inflation even during pilot errors or turbulence.
Intermediate/XC
Typical Range: 50-90 Pa
Cruise: 65-75 Pa
Collapse Threshold: ~28 Pa
Good stability with better performance. Requires some active piloting in strong conditions but maintains predictable behaviour.
Advanced/Sport
Typical Range: 40-80 Pa
Cruise: 55-65 Pa
Collapse Threshold: ~22 Pa
Higher performance with reduced stability margins. Requires active piloting and experience to manage in turbulent conditions.
Competition/Expert
Typical Range: 35-70 Pa
Cruise: 45-55 Pa
Collapse Threshold: ~18 Pa
Maximum performance with minimal margins. Demands constant attention and highly skilled piloting.
Machine Learning
Raw pressure data alone isn't enough. We're building machine learning models that learn to distinguish between pre-collapse pressure signatures and normal turbulence, pilot inputs, or benign thermal fluctuations.
The model architecture uses a two-stage approach:
Differential pressure + IMU data processed at high frequency. Detects asymmetric pressure events within 200ms. Provides the early warning signal.
Fast — <200ms latencyBarometric + GPS data confirms whether the spatial event is developing into a genuine collapse. Reduces false positives to operationally acceptable levels.
Confirms in 1–3 secondsNeural networks trained on intentional SIV collapses detect pre-collapse pressure signatures 2-3 seconds before loss of control.
Algorithms identify optimal thermal entry angles and circling techniques by correlating GPS track with pressure stability.
Automated risk assessment scores flights based on pressure margins, asymmetry events, and G-force exposure.
Personalised recommendations based on your flying patterns help identify areas for skill development.
Reference Guide
General guidelines for interpreting differential pressure readings. Actual values vary by wing design, EN class, loading, and conditions.
Important: These are general reference values. Your wing's specific characteristics, your flying weight, air density (altitude/temperature), and wing loading all affect absolute pressure values. Use ParaBaro's flight history to establish your wing's baseline pressures in calm conditions, then monitor deviations from that baseline.
The Dataset
No publicly available dataset contains synchronised internal canopy pressure data with GPS flight traces. The ParaBaro beta programme is building the world's first such dataset — and every hour of data makes the models smarter and more reliable.
Training requires thousands of hours of labelled flight data. This is why the beta programme exists — every flight uploaded by our 50 beta pilots feeds the model with each hour.
See What Your Wing Is Doing
Join 50 pilots building the world's first wing pressure dataset. Fly your normal hours, earn credits, and help make paragliding safer for everyone.