View source for UAS Search Patterns

Air searches are not a new concept - and they certainly are not exclusive to unmanned aircraft. Organizations such as the [[Example Manned Aircraft Search Patterns|Civil Air Patrol and the US Coast Guard have spent decades developing and refining techniques for conducting air searches]]. These methods have been used successfully all over the world, but are intended for manned aircraft operating over large areas and at altitudes above the typical small UAS flight. CDRP has developed modified search patterns and profiles that are based on proven techniques, but are adapted for the small UAS to utilize.  

These UAS Search Patterns are appropriate for searches for persons or material/objects (providing it is large enough to be detected). They should be utilized as part of a [[search strategy]] - all searches are more effective when all available resources are working together in a coordinated way.

==Search Methods and/or Patterns==
These search methods utilizing sUAS are guidelines - not hard and fast templates. The search manager and sUAS RPIC should collaborate on the effective usage of sUAS in the overall search effort, keeping in mind the strengths and weaknesses of the sUAS platforms and crews. 

===Hasty Search===
The Hasty Search is a amalgam of a variety of search tactics that are targeted at specific likely locations of the subject. The Hasty Search sounds unorganized, but the term comes from the fact that it is typically conducted first, with resources that are available immediately. 

In the application of sUAS in search, the Hasty Search tactic may include variety of searches that typically have a few common factors:
*The sortie is a [[Glossary#live_search|live search]].
*The sortie is flown manually - no automated flight plans or search patterns.
*The sortie is targeted by a search manager or team leader. "Go check that waterline, then all the stands of trees in the Division area." or "Check the bottom of the cliffs in Division B."

Hasty Searches (on the ground or in the air) are often effective, and should be directed at areas where there is a high chance of immediate success; the so-called low hanging fruit. In many scenarios, sUAS can search areas faster and safer than ground based searches, such as elevated structures, swift-water areas, ravines, and other areas that would require high-angle technical rescue. In cases where a child or at-risk adult has wandered or eloped, a Hasty Search should be conducted in the immediate vicinity of the subject's home, residence or care facility.  

'''Effective Use Scenarios:''' Unstructured initial searches of areas the subject is ''likely'' to be. Locations or areas where the sUAS can reach ''faster'' or ''safer'' than ground based search resources. 

'''Planning the Hasty Search:''' Altitude AGL varies - often flown very low, very close to obstacles and structures. sUAS crew should be mindful of regulations and risk to persons on the ground.

While the Hasty Search does not typically involve automated pre-planned flight, the sortie or sorties should be carefully planned with the search manager or search team leader. Experience, practice and intuition are important to selecting appropriate search targets and areas for immediate resources to cover. By operating under the direction of the search manager, the sUAS resource avoids duplication of effort and minimizes the risk of missing areas.

The aircraft should be operated at appropriate altitudes and distances to structures, terrain features and obstacles. Sensors appropriate for the target, conditions, and other factors should be employed.

'''Products of the Hasty Search:''' The main product of the Hasty is the initial, rapid, live search. In the event that a target is identified, the product delivered to the search manager should be the location or coordinates of the target as well as possible a photo or other supporting information. Coordinates should be in a format pre-identified with the search manager, search team leader, or planning section. Video recording, with positional information if available, should be made of the sortie, and should be examined when time permits. 

'''Integration of the Hasty Search:'''  The ad-hoc nature of the sUAS Hasty Search makes it relatively easy to integrate into other search resources' workflows. An embedded sUAS team can perform hasty search sorties as part of a ground search effort, supporting other search resources. For examples, a ground team working an area with elevated structures or ravines could utilize an embedded sUAS to conduct hasty searches of terrain or features that would otherwise be difficult or unsafe to reach.

===Parallel Track Search===
[[File:uas_grid_plan_sample1.png|thumb|Sample parallel track search flight path (green lines)]]The Parallel Track Search is the basic grid search - the sUAS is flown in a series of parallel 'tracks' or 'legs' covering an entire designated search area. sUAS Operators often refer to the Parallel Track Search as 'mowing the lawn' due to the resemblance of the search pattern to the path taken to mow the grass.

The Parallel Track Search can be used to cover a large or small area, but it typically best on relatively flat terrain focused on the subject's last known point (LKP) or other datum. The search pattern requires post flight analysis of the captured imagery to increase effectiveness and generate the highest possible probability of detection (POD).

This pattern is often employed by sUAS operators because it is the default pattern used by [[Glossary|commercial mapping software]]. These software programs will automatically generate a flight path appropriate for capturing [[Glossary|nadir images]] used to construct [[Glossary|ortho-mosaic photos]] or other [[Glossary|photogrammetry]] products. 

'''Effective Use Scenarios:'''  The Parallel Track Search is a deliberate, programmed flight pattern. It is best employed on flat or level terrain where tree cover is broken enough to allow sensors to see the ground. No sUAS sensors (with the possible exception of LIDAR) permit view through tree canopy and other overhead obstructions.

'''Planning the Parallel Track Search:''' Flight altitude is determined by a variety of factors, including: [[Glossary|sensor sweep width]], desired [[Glossary|Probability of Detection (POD)]], desired sweep [[Glossary|overlap]] or [[Glossary|sidelap]], and terrain and other obstacles.

Sensor sweep width is a variable of camera or sensor [[Glossary#angle_of_view|angle of view (AOV)]] and altitude. The higher the UA operates, the wider sweep width; the lower the altitude, the narrower the sweep. At the same time, the higher the UA (and therefore the wider sweep width) the resolution or [[Glossary|Ground Sample Distance]] is reduced. To gain more detail, the UA must be flown lower, while reducing sweep width of a single track, and resulting in higher flight times. This simple trade-off is the heart of the search problem using sUAS.

[[Glossary#Sensor_angle|Sensor angle]] will should be set based on a variety of factors. For processing ortho-mosaics, a nadir or near-nadir sensor orientation should be used. However, in situations where there is significant tree canopy, [[Glossary#oblique_images|oblique images]] may be more effective at an angle up to 45 degrees from vertical. The angled sensor orientation mitigates some canopy or overhead coverage; oblique images can see 'under' some obstruction. To maximize coverage using oblique images, planners should narrow the track spacing, resulting in higher overlap and reduced coverage. However, the tighter track spacing allows a specific point on the ground to be seen from at least two oblique angles. Planners should also remember that calculated GSD is only valid for nadir imagery. When oblique imagery is utilized, the GSD becomes higher closer to the horizon.

The Parallel Track Search is typically planned using automated software, typically [[Glossary#commercial_mapping_software|commercial mapping software]]; the parallel leg pattern is identical to the typical flight plan for mapping and photogrammetry. While mapping and photogrammetry may require a [[Glossary#sidelap|sidelap]] of 65% or greater, the sidelap in the Parallel Track Search could be considerably less (as low as 15%) to increase coverage per sortie. Planners should understand that a flight with overlap less than 60% may be sufficient for search purposes but will not be sufficient for building [[Glossary#ortho-mosaic_photos|ortho-mosaic]] or mapping products.

As overlap decreases, the amount of area than can be covered in a single [[Glossary#sortie|sortie]] will increase. Planners should understand that the flight altitude will also affect sweep width and therefore coverage on a single sortie, but as altitude increases, the [[Glossary#ground_sample_distance|ground sample distance]] or [[Glossary#resolution|resolution]] will decrease. In other words, the flight should be planned at the highest altitude at which an appropriate GSD can be achieved. This will maximize coverage while still providing appropriate sensor performance.

'''Products of the Parallel Track Search:''' The primary product of a Parallel Track Search is a list of potential targets or targets of interest. This list can be generated in one of two ways: [[Glossary#live_search|live search]] or [[Glossary#post-flight_search|post-flight search]]. Live search is less likely to result in a successful find, primarily due to environmental conditions including glare, poor screens, time pressure, and distraction. As a result of this, Parallel Track Search data should be reviewed using post flight search or 'the squint'. 

In addition to a target list, the search data can be processed into an ortho-mosaic or other GIS product(s) if there was sufficient coverage and overlap of nadir images.  

'''Integration of the Parallel Track Search:''' This search type can be utilized in conjunction with ground resources operating in the same [[Glossary#Division_area|division area]], or it can be performed before a ground team begins working that area. This [[Glossary#Leapfrog_tactic|leapfrog tactic]] allows the sUAS team to perform an initial search as well as produce near real-time maps and other information that can be provided to ground resources.

===Overlapping Grid Search===
<div style="float:right; padding-left: 10px;">
<gallery>
File:overlapping_grid_1.png|Sample overlapping grid search flight path, first pass (green lines)
File:overlapping_grid_2.png|Sample overlapping grid search flight path, second pass (green lines)
</gallery>
</div>
The Overlapping Grid Search is a modification of the Parallel Track Search (above).  When utilizing an oblique sensor angle, the Overlapping Grid provides additional views under canopy and other overhead concealment.

With most software, the Overlapping Grid is programmed as two separate sorties or flights at the same altitude and overlap settings, with the second track path offset by approximately 90 degrees. When the datasets from the two flights are combined, each point on the ground is visible from at least two (typically four) oblique angles - allowing for better coverage under trees, obstacles, building eaves, etc.  This search pattern is typically overkill for creating a simple orthomosaic product, but can still be processed in this way. 

'''Effective Use Scenarios:'''  The Overlapping Grid requires more time to complete, analyze and process, so it should be used only in situations that warrant the additional resources. It is best in scenarios where tree cover is heavy, and extended coverage is required to maximize the POD. Operators and search managers should understand that tree cover may defeat an airborne search (manned or unmanned) and while this search technique may increase the odds of detection, the additional resources required (time) may indicate that other search techniques are appropriate. 

'''Planning the Overlapping Grid Search:''' Flight altitude is determined by a variety of factors, including: [[Glossary|sensor sweep width]], desired [[Glossary|Probability of Detection (POD)]], desired sweep [[Glossary|overlap]] or [[Glossary|sidelap]], and terrain and other obstacles.  Each part of the search should be planned at the same altitude and sensor angle, and the two parts should be flown consecutively, with minimal time separation.

'''Products of the Overlapping Grid Search:''' The primary product of an Overlapping Grid Search is a list of potential targets or targets of interest. This list can be generated in one of two ways: [[Glossary#live_search|live search]] or [[Glossary#post-flight_search|post-flight search]]. Live search is less likely to result in a successful find, primarily due to environmental conditions including glare, poor screens, time pressure, and distraction. As a result of this, Overlapping Grid Search data should be reviewed using post flight search or 'the squint'. 

'''Integration of the Overlapping Grid Search:''' This search type can be utilized in conjunction with ground resources operating in the same [[Glossary#Division_area|division area]], or it can be performed before a ground team begins working that area. This [[Glossary#Leapfrog_tactic|leapfrog tactic]] allows the sUAS team to perform an initial search as well as produce near real-time maps and other information that can be provided to ground resources.

===Track Line Search===

===Creeping Line Search===

===Expanding Square Search===

===Sector Search===

===Shoreline Search===

===Barrier Search===