Request a briefing

Detection

How to detect UAVs before they reach the site: radar, optics, RF and acoustic sensors

Detection is not a moment but a chain of confirmations. Radar sees movement, optics see shape, RF sees the control link. Individually each sensor is wrong often enough; together they produce a picture you can act on.

October 14, 20268 minSensors
Optoelectronic UAV search system AEROTEGIS
ContentsFour physicists, four different answersWhy does one sensor always lie?What does range consist of?Search and support are different tasksWhat happens after detectionFive questions for the tender

In tenders, detection is often described with one number - “range up to 5 km.” In practice, the system can see a target at a distance of 5 km and not be able to identify it, issue dozens of false alarms per day, or lose a low-flying drone in the background of construction. Therefore, design begins not with range, but with the question: what targets, under what conditions and with what reaction time need to be detected.

Four physicists, four different answers

ChannelStrengthLimitation
RadarRange, all-weather, speed measurementDoes not distinguish target type, depends on RCS and background
Optics and IRIdentification, visual recording, outcome evaluationRequires clear line of sight, weather dependent
Radio intelligenceControl channel and video detectionDoes not see autonomous and fiber optic drones
AcousticsPassivity, work on low-flying targetsShort range, city noise

None of these channels provide both range, accuracy and weather resistance. Therefore, in a detection system they do not compete, but complement each other. Radar is the first to detect movement, optics confirm the type, radio reconnaissance specifies the control channel, acoustics closes blind spots under the canopy and behind buildings.

Why does one sensor always lie?

The radar sees birds, hot air balloons, reflections from buildings and wind turbine blades. The optics mistakes a package attached to a tree for a drone. Radio intelligence hears other people's channels and household routers. If the system cannot classify targets, the operator receives a stream of false positives and stops trusting it.

The practical criterion is not “how many targets it sees,” but how many false alarms per day for a given detection level. This parameter should be asked before purchase and checked on site, not in the showroom.

What does range consist of?

For a radar, the range depends on the effective scattering area of ​​the target, flight altitude, background and processing algorithm. For optics - on target size, contrast, atmospheric visibility and focal length. Therefore, “5 km” is the detection range of a specific class of targets under specific conditions, and not a universal figure.

5 km
Detection
small UAV above the building line
2 km
Identification
optical-electronic confirmation
≤ 3 s
Search cycle
full 360° view
360°
Sector of review
azimuth and elevation

Search and support are different tasks

Search OES has a wide field of view and a fast scanning cycle. Tracking system - narrow field, long focus and high accuracy of angular measurements. Trying to solve both problems with one block always leads to a compromise: either a wide view or long-range recognition.

Hence, there are two architectural solutions: integrated OES “search + tracking” in one building and separate system, where the search and tracking blocks are separated. The first deploys faster, the second gives better range and flexibility of placement on a complex object.

Integrated optical-electronic search and tracking system
Integrated OES: broad search and narrow support in one package. Detection range is up to 5 km, identification range is up to 2 km.

What happens after detection

  1. Forming a route and assigning a threat level.
  2. Guidance of EOS tracking and visual identification.
  3. Video recording and logging of operator actions.
  4. Transfer of target designation to the means of influence.
  5. Evaluate the result and return to the search.

Each step takes time. If the system detects a target at 5 km and the approach lasts two minutes, the operator has time to confirm and make a decision. If the target appears at 500 m, there is almost no time - and then the automation, pre-configured to certain rules, decides.

Five questions for the tender

  • Probability of detection with RCS of 0.01 m² and visibility of 10 km.
  • How many false alarms per day at an object with a given background.
  • Range of identification, not just detection.
  • Interfaces and protocols for integration with existing security systems.
  • Who and how conducts full-scale tests at the facility.
Is it possible to detect a drone without radar?
Yes, optics, radio intelligence or acoustics. But each channel has limitations: optics depend on the weather and line of sight, radio reconnaissance does not see autonomous and fiber optic drones, acoustics work at a short distance.
What is EPR 0.01 m²?
Typical effective dispersion area of ​​a small quadcopter. For comparison: a bird has an EPR of the order of 0.001–0.01 m², a large aircraft has tens of square meters.
Why is the identification range less than the detection range?
To distinguish the type of target, you need to collect a sufficient number of pixels on the object. At a long distance, the target occupies a few pixels - this is enough for alarm, but not for confident recognition.
Is urban development a problem?
Yes. Buildings create radar shadows and blind spots. Therefore, at complex facilities, several positions and a distributed optical-electronic network are used.
Is it possible to combine the system with an existing radar?
Yes, if open integration protocols are provided. This is a question to ask before purchasing: a closed system may require replacing existing equipment.

Start with the site, not the hardware.

Tell us what you are protecting, the approximate perimeter and the threats you are seeing. We will come back with a layered configuration and a deployment sketch.