Airffic InstituteAERIAL PHYSICAL AI

Aerial Physical AI Institute

Sensors: how a drone sees the world

No animal has just one sense. Drones, too, combine eyes with different strengths.

Short answer: Cameras are cheap and rich in detail but struggle in darkness and fog. Radar handles weather but not fine shapes. LiDAR measures distance precisely but is heavy and costly. That is why combining sensors (sensor fusion) is the winning approach.

Updated: · 3 min readHow we researched →

Sir Airffic GoOnce you know the sensors, you can see at a glance what each company is betting on.

Six kinds of eyes and senses

Cameras capture light like human eyes. They are cheap, light and see colour and text, so they are the most widely used.

LiDAR fires laser pulses and times the reflections to measure distance precisely. It works in the dark, but weight and cost are challenges.

Radar uses radio waves. It spots distant objects in rain, fog and darkness, so it leads in collision avoidance and drone detection.

Thermal cameras see temperature differences — ideal for night search and for finding hot spots on roofs and power lines.

The IMU (measures acceleration and rotation) and GNSS (satellite positioning such as GPS) are the drone's inner senses. Only combined with outward-looking sensors do they give precise flight.

SensorStrengthsWeaknessesMain uses
Camera (visible)Cheap, light, rich detailDarkness, fog, glareRecognition, SLAM, inspection
Stereo / depth cameraDistance at short rangeInaccurate far awayObstacle avoidance
LiDARPrecise range, works in darkHeavy, costly, weaker in rainSurvey, thin-wire detection
mmWave radarRain, fog, night, long rangeFine shape detailDetect-and-avoid, drone detection
ThermalSees heat, strong at nightLower resolutionSearch & rescue, asset inspection
IMU + GNSSOwn motion and positionGNSS fails indoors or when jammedAttitude control, navigation

Why combining sensors decides the game

Every sensor has a weakness. Rely on one, and the drone goes blind the moment that weakness appears.

So a camera identifies what something is, radar confirms how far away it is, and the IMU fills in how the drone moved. This is sensor fusion.

Companies good at fusion build aircraft that keep working when weather and location change. Not stopping is the most valuable feature of a commercial drone.

Skies without GNSS raised the value of sensors

GNSS signals do not reach under bridges, in tunnels or inside warehouses. On battlefields, deliberate jamming became routine.

Demand for flying with cameras and an IMU alone (visual navigation) rose sharply after 2022.

This is not only military. The same technology helps indoor inspection and urban delivery.

Sir Airffic GoCan it fly when radio fails? That is the best test of whether autonomy is real.

What is happening in the sensor market

Camera parts mass-produced for smartphones made drone eyes dramatically cheaper.

Radar and LiDAR are shrinking and getting cheaper thanks to automotive volumes, so they now fit on drones.

Many sensors therefore evolve as by-products of huge non-drone markets. Drone specialists must differentiate on how they combine sensors and on software, not on the parts.

Frequently asked questions

Which sensors do drones use for obstacle avoidance?

Stereo or depth cameras at short range, radar at long range and in bad weather, and LiDAR for thin objects such as wires. Most aircraft combine several.

What is the difference between LiDAR and radar?

LiDAR uses laser light to measure distance finely and build detailed 3D shapes. Radar uses radio waves; it is coarser but finds distant objects in rain, fog and darkness.