Drone Guide Avalanche Search

Avalanches create one of the most time-critical and difficult search environments faced by mountain rescue teams. Snow can move at high speed, carry people hundreds of metres, bury victims beneath metres of debris and leave a search area that is unstable, cold and difficult to access. Rescuers may also face continuing avalanche danger, poor visibility, strong winds and steep terrain while attempting to locate missing people as quickly as possible.

Drones can provide mountain rescue teams with an additional aerial search capability without immediately exposing personnel to every part of the avalanche debris field. Equipped with combinations of RGB cameras, thermal imaging, mapping sensors, spotlights, loudspeakers and, where technically and operationally appropriate, specialised radio or avalanche-search equipment, drones can rapidly survey large areas and provide information to incident commanders and rescue specialists.

Their role should be understood carefully. A drone does not replace avalanche transceivers, organised probing, rescue dogs, Recco-type detection systems, ground search teams, helicopters or professional avalanche-rescue procedures. Snow can prevent many sensors from detecting a completely buried person, and thermal cameras in particular cannot simply see through deep snow.

The strongest approach is therefore to integrate drones into a wider rescue system in which aerial reconnaissance, conventional avalanche-search methods, rescue teams and crewed aviation work together.

The Challenge of Avalanche Search

An avalanche site can be significantly larger and more complicated than it initially appears. Snow may have travelled hundreds of metres from the release zone, carrying people, equipment, trees and debris with it. Victims can be located close to the surface, partially buried or completely buried.

Search teams need to determine the likely avalanche path, identify where missing people were last seen, establish possible burial areas and look for visual clues such as skis, clothing, backpacks or other equipment.

At the same time, rescuers must consider their own safety. Secondary avalanches remain a serious concern, particularly when unstable snow remains above the search area.

A drone provides a way to obtain an immediate overview before or while ground resources are being organised.

Rapid Aerial Reconnaissance

One of the most valuable roles for a drone is rapid reconnaissance.

Shortly after arriving at an avalanche incident, a drone can potentially be launched to survey the debris field and surrounding terrain. Live video can help rescuers understand the size of the avalanche, the direction of snow movement and the accessibility of different areas.

This aerial perspective can be difficult to obtain from ground level.

The drone can also help identify terrain features that may have influenced where victims or equipment were carried.

Information can be streamed to the incident commander, allowing search resources to be coordinated using a common view of the site.

Mapping the Avalanche Debris Field

Mapping the avalanche is an important early task.

RGB imagery can be used to document the boundaries of the debris field, while photogrammetry or LiDAR-equipped drones may provide more detailed terrain information when required.

The resulting map can help divide a large avalanche into manageable search sectors.

Search teams can then record which areas have been investigated and which still require attention.

Mapping also creates a useful reference for rescue coordination. Instead of describing a location verbally as “near the large rock” or “below the trees,” teams can work from georeferenced locations within a shared map.

However, a mapped avalanche surface does not reveal everything beneath the snow. Mapping should support the rescue operation rather than create false confidence that hidden areas have been cleared.

RGB Cameras for Avalanche Search

High-resolution RGB cameras are among the most useful drone sensors during avalanche response.

Aerial imagery can reveal partially buried people and equipment that may be difficult to see from ground level.

Rescuers can look for skis, poles, backpacks, helmets, clothing and other objects on the snow surface.

These clues can be important because equipment may indicate the path travelled by a missing person.

Modern zoom cameras can allow operators to examine suspicious objects without immediately moving rescuers across potentially unstable snow.

However, visual identification remains dependent on lighting, visibility, snow conditions and the amount of an object exposed above the surface.

Searching for Surface Clues

Avalanche victims are not always completely buried.

A hand, boot, helmet or piece of clothing may remain visible.

The aerial perspective provided by a drone can make some of these clues easier to identify.

Operators can systematically scan the debris rather than relying only on observations from the edge of the avalanche.

High-resolution photographs can also be reviewed by additional personnel away from the immediate rescue site.

AI-based image analysis may eventually help highlight unusual shapes or colours within large image sets, but any automated detection should be treated as a candidate observation requiring human review.

A missed visual clue cannot be interpreted as evidence that no victim is present.

Thermal Imaging

Thermal cameras are frequently considered for search and rescue because they detect differences in surface infrared radiation associated with temperature.

If a person is exposed or only partly covered, thermal imaging may help identify them, particularly when there is useful temperature contrast between the person and surrounding environment.

Thermal cameras can also assist with locating rescuers, vehicles or other warm objects during low-light operations.

However, thermal imaging has a major limitation in avalanche rescue:

thermal cameras do not reliably see through significant snow cover.

Snow acts as an insulating layer and blocks the direct infrared radiation from a buried person.

A thermal camera should therefore not be treated as a detector capable of locating every buried avalanche victim.

Partially Buried Victims

Thermal imaging may be particularly useful when part of a victim remains exposed.

A face, hand or section of clothing warmed by the body could potentially create detectable thermal contrast.

The effectiveness will depend on burial conditions, temperature, wind, time since burial and sensor performance.

Operators should inspect thermal anomalies carefully using the accompanying RGB camera.

A warm spot is not automatically a person.

Sun-heated rocks, equipment, vegetation or rescue personnel can also produce thermal signatures.

Thermal imagery therefore provides additional information rather than definitive identification.

Thermal Anomalies on Snow

There may be situations where heat transfer influences the snow surface, but this should be interpreted very cautiously.

Snow depth, density, temperature and environmental conditions strongly influence heat transfer.

A deeply buried person may produce no useful surface thermal signature.

Conversely, a temperature difference on the snow surface may have an unrelated environmental explanation.

For operational planning, rescuers should assume that a conventional airborne thermal camera cannot reliably detect a person through deep avalanche debris.

This prevents thermal imagery from creating a dangerous false negative.

Avalanche Transceivers

Avalanche transceivers are one of the principal technologies used to locate buried people who are carrying compatible transmitting devices.

A drone-based system could potentially support certain forms of radio-based search if specifically designed, tested and approved for that purpose.

However, this is technically more complex than simply attaching an antenna to a drone.

The aircraft’s motors, electronics, communications systems and power distribution can introduce electromagnetic interference.

Altitude above the snow also changes the relationship between the receiver and buried transmitter.

Any drone-based avalanche-transceiver solution therefore needs dedicated engineering and operational validation.

It should complement established transceiver search procedures rather than replace them without strong evidence.

Radio-Frequency Detection

More broadly, drones can carry radio-frequency sensing equipment capable of detecting certain electronic transmissions.

This creates potential opportunities for locating devices carried by missing people.

However, detecting a radio signal is different from locating a person precisely.

Signal strength changes because of distance, antenna orientation, terrain, snow and reflections.

A detected signal may help define an area for further investigation, but professional rescue procedures remain necessary to determine the victim’s location.

The absence of a detected signal also does not establish that nobody is present.

Mobile-Phone Detection

Missing skiers or mountaineers may carry mobile phones.

In some circumstances, authorised search systems may use cellular or other radio information as an additional search input.

Drone-mounted RF equipment could potentially help map signal observations over difficult terrain.

However, mobile-phone detection has significant technical, legal and privacy considerations.

A phone may be switched off, damaged, out of battery or outside network coverage.

Snow and terrain may also affect radio propagation.

A cellular signal therefore does not equal a confirmed victim location, while failure to detect a phone does not establish that the search area is empty.

RECCO and Specialist Detection Technologies

Some outdoor clothing and equipment incorporates passive rescue reflectors designed to be located by compatible specialist detectors.

Airborne implementations of specialist avalanche-search technologies can potentially extend the area that can be surveyed from above.

Where such equipment is available, it can provide another search layer alongside avalanche transceivers, dogs and probing.

However, compatibility is essential. A detector can only identify technology it has been designed to detect.

A negative result therefore cannot exclude the presence of a buried person who is not carrying compatible equipment or whose position cannot be detected under the specific conditions.

Search Dogs and Drones

Avalanche rescue dogs remain an important capability because they can search for human scent without requiring the victim to carry electronic equipment.

Drones and dogs can complement one another.

The drone can provide an overview, map the avalanche and identify visual clues while dog teams search selected areas on the ground.

Drone operations should be coordinated carefully around search dogs.

Aircraft noise and rotor wash may distract animals or disturb scent conditions.

Incident commanders should therefore determine when and where aerial operations are appropriate rather than flying continuously over active dog teams.

Probe Searches

Physical probing remains important when rescuers need to search snow for buried victims.

Drones cannot physically confirm a burial location in the way a probe can.

Instead, aerial information can help organise the wider operation and document areas being searched.

Digital mapping can potentially record search sectors, clues and confirmed locations.

This combination of aerial situational awareness and physical ground search is considerably more useful than treating the drone as an independent rescue system.

Search-Sector Management

Large avalanche fields can be divided into search sectors.

Drone imagery can help define these areas and provide a shared map for rescue teams.

Locations of clues can be marked digitally.

Areas searched by different teams can also be recorded.

This helps incident commanders understand progress and reduces the risk that difficult sections are unintentionally overlooked.

However, software-generated coverage should reflect what search method was actually used. Flying a camera over an area is not equivalent to clearing it using avalanche transceiver search, probing or other specialist techniques.

Last-Seen-Point Analysis

Information about where a missing person was last observed is extremely valuable.

A drone can map that location in relation to the avalanche path.

Rescuers can then examine terrain between the last-seen point and the debris field.

Objects identified from the air can be plotted on the same map.

This creates a spatial picture of the incident.

However, trajectory assumptions should remain cautious. Avalanche movement is chaotic, and equipment may become separated from a person.

A recovered ski or backpack is an important clue rather than definitive evidence of the victim’s exact location.

Terrain Traps

Terrain strongly influences avalanche consequences.

Depressions, gullies, trees and other terrain features can affect how snow accumulates.

Aerial mapping provides rescuers with a clearer view of the overall terrain.

This information can support professional avalanche specialists when assessing the incident area.

However, a drone image alone should not be used to declare terrain safe.

Snow stability requires appropriate avalanche expertise and observations.

Secondary Avalanche Risk

One of the strongest arguments for using drones is reducing unnecessary human exposure.

If unstable snow remains above the avalanche site, sending rescuers immediately across every part of the debris field may expose them to additional danger.

A drone can conduct initial reconnaissance from a safer position.

It can inspect surrounding slopes visually and provide information to avalanche professionals.

However, the drone cannot independently determine that a slope is safe.

Decisions about responder access should remain with qualified rescue and avalanche personnel.

Search and Rescue Helicopters

Helicopters remain extremely important in mountain rescue.

They can transport rescue teams, evacuate injured people and provide aerial search capability over large areas.

Drones should therefore be integrated carefully with crewed aviation.

Uncoordinated drone operations near rescue helicopters can create a serious collision hazard.

Where helicopters are operating, drone flights should follow the incident’s airspace coordination procedures.

Crewed rescue aviation takes priority.

Drone Deployment Before Helicopter Arrival

In some incidents, a small rescue drone may reach the avalanche site before a helicopter can arrive.

This can provide early imagery to rescue coordinators.

The information may help arriving teams understand the scale of the incident and identify visible clues.

Once crewed aircraft approach, drone operations may need to stop or move according to established coordination procedures.

The goal is not competition between drones and helicopters.

It is to provide the incident commander with the most useful combination of available assets.

Poor Visibility

Mountain weather can change rapidly.

Cloud, snowfall and blowing snow can make visual searching difficult.

LiDAR and some radar technologies can operate independently of visible light, but their ability to map terrain does not mean they can identify a person beneath snow.

Thermal cameras may still operate in darkness but can also be affected by environmental conditions.

The aircraft itself must remain capable of safe operation.

A sensor’s theoretical capability is not useful if wind, icing or visibility prevents the drone from being flown safely.

Night Operations

Avalanche incidents can continue after sunset.

Drones equipped with thermal cameras, low-light cameras and searchlights can support night operations.

A searchlight can illuminate exposed objects for RGB cameras and ground teams.

Thermal imaging can help locate exposed or partially exposed people.

However, snow can produce glare under strong illumination.

Operators should adjust camera exposure and lighting direction carefully.

Night operations may also require additional aviation approvals or procedures depending on the jurisdiction.

Searchlights

A drone-mounted searchlight can illuminate specific areas without requiring rescuers to move immediately onto the debris field.

This can be particularly useful around cliffs, trees or avalanche edges.

The light can also help rescuers navigate once they enter the area.

However, illumination should not be treated as evidence that a route is safe.

Snow bridges, unstable debris and steep drops may remain difficult to identify.

Ground personnel still need appropriate mountain-rescue procedures.

Loudspeaker Payloads

Loudspeakers can provide another useful capability where a missing person is conscious and not completely buried.

The drone may broadcast instructions or ask the person to signal visually or audibly.

This could be useful when searching large mountain areas surrounding an avalanche.

However, wind and terrain can significantly reduce intelligibility.

A loudspeaker broadcast does not confirm that a person heard or understood the message.

It should therefore complement rather than replace other communication and search methods.

High-Resolution Zoom Cameras

Optical zoom allows operators to inspect suspicious areas while keeping the aircraft at a safer distance.

A small dark object on the snow can be examined without immediately sending a rescuer toward it.

Zoom cameras are particularly useful around cliffs or unstable slopes.

However, magnification also increases the effect of aircraft movement.

A stabilised gimbal is important.

Operators should capture still images where useful because they can sometimes reveal more detail than live compressed video.

Mapping with Photogrammetry

Overlapping drone photographs can create an orthomosaic and three-dimensional model of the avalanche.

This provides a detailed record of the incident.

The model can support search-sector planning and later avalanche analysis.

However, photogrammetric processing takes time.

During the most urgent rescue phase, live video and rapidly generated maps may be more useful than a highly detailed final model.

The mapping workflow should therefore match the urgency of the operation.

LiDAR for Avalanche Mapping

LiDAR can rapidly measure the three-dimensional surface of avalanche debris.

This may help document debris depth, terrain and avalanche morphology when combined with pre-event terrain data.

LiDAR can also work without visible surface texture.

However, ordinary aerial LiDAR does not see a buried person through metres of snow.

Its role is primarily terrain and surface mapping.

It should not be described as a victim detector unless paired with another specifically validated technology.

Pre- and Post-Avalanche Terrain Models

If high-quality terrain data exists from before an avalanche, a post-event drone survey can be compared with it.

The difference may show where snow accumulated or where terrain was exposed.

This can be valuable for avalanche research and recovery analysis.

During an active rescue, however, the priority remains locating people.

Detailed volumetric modelling is generally secondary to time-critical search information.

After the incident, the same drone dataset can provide substantial value for analysis and documentation.

GNSS and Mapping

GNSS allows observations from the drone to be linked to coordinates.

A visual clue can be marked on a map and passed to a ground team.

RTK or other high-accuracy positioning can improve the precision of mapping products.

However, a coordinate derived from an oblique camera image may not have the same accuracy as the drone’s own GNSS position.

Operators should distinguish approximate target coordinates from surveyed positions.

Ground teams should expect to search around candidate locations rather than assume centimetre-level victim positioning.

Mountain GNSS Challenges

Mountain terrain can block portions of the sky and create challenging GNSS geometry.

Cliffs and steep valleys may reduce positioning quality.

Snow-covered environments can also contain few visually distinctive features for optical navigation systems.

Drone operators should therefore monitor navigation quality carefully.

The ability to launch does not guarantee that all positioning systems will perform equally throughout the search area.

Wind

Strong mountain winds can significantly affect drone operations.

Avalanche sites may experience turbulence around ridges and cliffs.

A drone flying slowly while searching can be particularly affected by gusts.

Wind also reduces battery endurance because the aircraft uses more power to maintain position.

Mission planning should therefore preserve sufficient energy reserve.

A search should not continue until the battery is nearly exhausted simply because an area remains unchecked.

Cold Temperatures

Cold weather reduces battery performance.

This is particularly important during avalanche rescue.

Batteries may need to be kept warm before use.

Flight endurance should be estimated conservatively.

The aircraft, camera gimbal and sensors may also have minimum operating temperatures.

Professional rescue organisations should validate their drone systems under realistic winter conditions rather than relying solely on room-temperature endurance specifications.

Snowfall and Moisture

Falling snow can interfere with cameras and LiDAR.

Snowflakes close to the sensor may appear as large visual objects or unwanted laser returns.

Moisture can also accumulate on lenses.

The aircraft’s environmental protection rating should be understood.

Not every commercial drone is designed for operation in snowfall.

Safe operating limits should not be exceeded because the mission is urgent.

Icing

Icing represents a serious risk for drones.

Ice accumulation on propellers can alter aerodynamic performance rapidly.

Sensors and airframes may also become affected.

A drone should not be assumed capable of flying safely in icing conditions simply because it can operate at low temperature.

Operational decisions should follow manufacturer limits and rescue-organisation procedures.

Battery Management

Avalanche response may require repeated flights.

A battery-management system should therefore be part of the rescue plan.

Charged batteries can be rotated while others are warmed and recharged.

Portable power systems may support remote operations.

Flight logs should track battery condition.

Cold-weather degradation can be particularly dangerous because voltage may fall more rapidly under load.

Adequate reserve is essential.

Drone-in-a-Box and Mountain Monitoring

In the future, automated drone stations could be located at ski resorts, mountain rescue bases or high-risk transport corridors.

Following an incident alert, a drone could potentially launch and provide initial reconnaissance before rescue teams arrive.

The aircraft might map the avalanche boundary and transmit imagery to the rescue centre.

However, fully autonomous mountain operations are challenging because of weather, snow accumulation, icing, communications and rapidly changing aviation activity.

Human rescue coordination will remain essential.

Ski Resorts

Ski resorts represent a potential environment for integrated drone rescue systems.

A drone stationed near a patrol base could provide rapid aerial reconnaissance following an avalanche report.

Resort maps, lift infrastructure and known ski routes could be integrated into the operational software.

However, resort airspace may become extremely busy during an emergency.

Helicopters, ski patrol and rescue teams may all be operating.

Drone deployment therefore requires established procedures rather than ad-hoc flying.

Mountain Roads and Railways

Avalanches also affect roads and railway lines.

Drones can survey blocked transport corridors and determine the visible extent of snow deposition.

Thermal and RGB cameras may assist if vehicles or people are involved.

LiDAR can document debris geometry.

However, a road appearing clear from the air does not mean it is safe to reopen.

Further avalanche risk and structural conditions require professional assessment.

AI-Assisted Visual Search

AI computer vision can assist operators by analysing imagery for candidate objects.

Algorithms may search for people, clothing, backpacks, skis or unusual colour patterns.

This can be useful when hundreds of images have been collected.

However, snow environments are difficult for automated detection.

Shadows, rocks and vegetation can create false positives.

A completely buried victim may provide no visual feature for AI to identify.

AI should therefore prioritise observations for human review rather than declare an area clear.

AI and Thermal Search

AI can also analyse thermal imagery.

Software may highlight warmer regions or track candidate objects between frames.

This can reduce operator workload.

However, the limitations of the thermal sensor remain unchanged.

AI cannot recover thermal information that never reaches the camera because the person is insulated beneath deep snow.

A sophisticated algorithm does not turn a thermal camera into a through-snow imaging system.

Multi-Sensor Search

The strongest drone systems may combine several sensors.

A single aircraft could carry an RGB zoom camera and thermal sensor, while another specialised platform carries radio-detection equipment.

Mapping software could combine all observations.

A visual clue, radio signal and terrain feature could then be displayed on one incident map.

This does not mean that multiple weak observations automatically confirm a victim.

Instead, sensor fusion can help rescue specialists decide where to concentrate proven search methods.

Data Fusion

Avalanche response involves information from many sources.

These may include witness statements, last-seen positions, avalanche transceiver observations, drone imagery, dog searches, probing, helicopter observations and mobile communications.

A GIS-based incident map can combine this information.

Drones are particularly useful because their observations can be spatially referenced.

However, each layer should indicate its confidence and source.

An approximate visual location should not appear identical to a confirmed probe strike.

Communications Relay

Mountain terrain can block radio communications between rescue teams.

Some drones can potentially act as temporary communication relays.

Elevating a radio node above terrain may improve line of sight.

This could help teams working in valleys or behind ridges.

However, communications relay and search operations may compete for aircraft endurance.

Dedicated systems may therefore be more appropriate for large incidents.

Search Progress Mapping

Digital maps can help record rescue progress.

Search sectors can be marked according to the method used.

For example, an area visually inspected by drone should be distinguished from one searched with transceivers or probes.

This prevents the map from overstating search confidence.

As teams report their progress, incident commanders gain a clearer understanding of remaining priorities.

Drone imagery can provide the base map supporting this coordination.

Privacy and Data Management

Avalanche rescue imagery may contain identifiable victims and rescuers.

Data should therefore be handled appropriately.

Live video may also capture private activities around ski areas or nearby properties.

Emergency use does not eliminate the need for responsible data management.

Access should be limited to personnel who need the information, particularly where imagery shows injured or deceased people.

Training

Drone avalanche search should be practised before an emergency occurs.

Pilots need experience operating over snow because the environment can be visually deceptive.

Rescue teams should understand what each sensor can and cannot detect.

Joint exercises can establish procedures for drone deployment, helicopter coordination, search-sector mapping and data sharing.

Training is particularly important for thermal imaging.

Teams should see first-hand how quickly snow can hide a person’s thermal signature.

Payload Selection

A drone intended for avalanche search should be selected around the actual rescue requirement.

Important considerations include cold-weather capability, wind resistance, flight endurance, high-resolution RGB zoom, thermal imaging, positioning, searchlight compatibility, communications range and rapid deployment.

Specialist RF or avalanche-detection payloads may provide additional capabilities where they have been properly validated.

Payload weight matters because heavier sensors reduce endurance.

In many cases, a reliable dual RGB/thermal camera with strong zoom and good flight performance may provide more operational value than an overloaded aircraft carrying numerous sensors.

Benefits and Limitations

The principal advantage of drones in avalanche search is speed of aerial access. They can provide rescuers with an immediate overview of the avalanche, search for visible or partially buried victims, identify equipment, map the debris field and reduce unnecessary exposure to unstable terrain.

Thermal cameras, zoom cameras, searchlights, loudspeakers and specialist detection equipment can expand these capabilities.

However, the limitations are equally important. A conventional thermal camera cannot reliably see through deep snow. LiDAR maps the snow surface rather than locating buried people. Visual cameras cannot detect a completely hidden victim. Radio-based systems only work when suitable signals or compatible equipment are present.

A negative drone search therefore does not mean that nobody is buried.

The drone should be considered an additional search layer within a professionally managed avalanche-rescue operation.

The Future of Drones in Avalanche Search

Future avalanche-rescue drones are likely to combine more sensors and greater autonomy.

AI may automatically map avalanche boundaries, identify exposed equipment and highlight candidate thermal signatures. Improved radio-detection systems could potentially assist with locating compatible devices. Automated flight planning may divide large avalanche fields into systematic search sectors.

Drone-in-a-Box systems could provide rapid initial reconnaissance around ski resorts and mountain transport corridors. Long-endurance platforms may provide communications relay while smaller drones perform close search.

The greatest development, however, is likely to come from better integration rather than any single new sensor.

A future rescue workflow could operate as:

avalanche alert → rescue coordination and airspace assessment → rapid drone deployment → avalanche-boundary mapping → RGB and thermal search → candidate visual or sensor observations georeferenced → integration with witness information and last-seen point → specialist avalanche-transceiver, detector, dog and probe searches → ground-team verification → medical rescue and evacuation → post-event LiDAR or photogrammetric mapping → professional incident review.

Throughout this process, drone observations would support the rescue specialists rather than independently determine whether a victim is present or whether an area is safe.

Conclusion

Drones can provide a powerful additional capability for avalanche search and rescue by giving responders rapid access to an aerial view of dangerous and difficult mountain terrain.

High-resolution RGB cameras can identify exposed victims and equipment. Thermal sensors can assist with exposed or partially buried people. Mapping systems can define the avalanche debris field and organise search sectors. Searchlights, loudspeakers and specialist detection payloads can provide further support under appropriate conditions.

Their greatest benefit may be the ability to gather information while reducing unnecessary exposure of rescuers to unstable terrain.

At the same time, the limitations of drone sensors must be clearly understood. Snow can completely hide a victim from visual and thermal cameras, and a negative aerial search cannot establish that an avalanche field is clear.

The strongest avalanche-response capability therefore combines drones, avalanche transceivers, specialist detection technologies, rescue dogs, probing, professional mountain-rescue teams, medical response and coordinated crewed aviation.

Used in this way, drones become an additional set of eyes and sensors above the avalanche—helping rescue teams understand the scene faster, direct resources more effectively and reach people in need as quickly and safely as possible.