AUTONOMOUS APPROACH AND LANDING
CAPABILITY (AALC) DEMONSTRATION
Delivery Order 0018: Opportune Landing Site (OLS) Software Field
Demonstration and Validation of Capability to Identify Landing
Sites and Low Incidence of False Positives
Carol Ventresca, Victoria M. Althoff, Kenneth R. Eizenga, and Capt Justin R. Rufa
Approved for public release; distribution unlimited.
See additional restrictions described on inside pages
AIR FORCE RESEARCH LABORATORY
AIR VEHICLES DIRECTORATE
WRIGHT-PATTERSON AIR FORCE BASE, OH 45433-7542
AIR FORCE MATERIEL COMMAND
UNITED STATES AIR FORCE
NOTICE AND SIGNATURE PAGE
Using Government drawings, specifications, or other data included in this document for any
purpose other than Government procurement does not in any way obligate the U.S. Government.
The fact that the Government formulated or supplied the drawings, specifications, or other data
does not license the holder or any other person or corporation; or convey any rights or permission to
manufacture, use, or sell any patented invention that may relate to them.
This report was cleared for public release by the USAF 88 th Air Base Wing (88 ABW) Public Affairs Office
(PAO) and is available to the general public, including foreign nationals. Copies may be obtained from the
Defense Technical Information Center (DTIC) (http://www.dtic.mil).
AFRL-RB-WP-TR-2008-3163 HAS BEEN REVIEWED AND IS APPROVED FOR PUBLICATION IN
ACCORDANCE WITH ASSIGNED DISTRIBUTION STATEMENT.
JUSTIN R. RUFA, Captain, USAF
Control Systems Development and
DANIEL B. THOMPSON
Control Systems Development and
JEFFREY C. TROMP
Senior Technical Advisor
Control Systems Development and
This report is published in the interest of scientific and technical information exchange, and its
publication does not constitute the Government’s approval or disapproval of its ideas or findings.
REPORT DOCUMENTATION PAGE
OMB No. 0704-0188
The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, searching existing data
sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of
information, including suggestions for reducing this burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a
collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std. Z39-18
Table of Contents
List of Figures.iv
List of Tables.vii
1. Executive Summary.1
2.1 Summary of Approach.2
2.2 Identifying Candidate OLSs.3
3. Methods, Assumptions, and Procedures.4
4. Results and Discussion.6
List of Acronyms, Abbreviations, and Symbols.37
List of Figures
Figure 1. Demonstration Area within St. Clair County.3
Figure 2. B1 Shown on Map.7
Figure 3. B1.7
Figure 4. B1 Another view.8
Figure 5. B1 Software Output.8
Figure 6. B2 Shown on Map.8
Figure 7. B2 (a) .8
Figure 8. B2 (b) Another View.8
Figure 9. B2 (c) From Another Direction.8
Figure 10. B2 (d) Acceptable Approach.9
Figure 11. B2 (e).9
Figure 12. B2 (f).9
Figure 13. B2 Software Output.9
Figure 14. B3 Shown on Map.9
Figure 15. B3.9
Figure 16. B3 Another View.10
Figure 17. Orthophotoquad of B3.10
Figure 18. B3. Software Output.10
Figure 19. Bl, and J15 As Shown on Map.10
Figure 20. J15.10
Figure 21. Orthophotoquad of J15.11
Figure 22. J15 Software Output.11
Figure 23. JX1 Shown on Map.11
Figure 24. JX1.11
Figure 25. JX1 From Another Direction.11
Figure 26. JX1 Another View.11
Figure 27. J8 and JX2.12
Figure 28. J8.12
Figure 29. J8 A Closer View.12
Figure 30. J8 and JX2.12
Figure 31. JX2 (a).13
Figure 32. JX2 (b) .13
Figure 33. JX2 (c).13
Figure 34. JX2 Repeat of Orthophotoquad.13
Figure 35. JX2 (d).13
Figure 36. J8.13
Figure 37. Map Showing Location of J7.14
Figure 38. J7.14
Figure 39. J7 Another View.14
Figure 40. J7 A Different Perspective.14
Figure 41. Orthophotoquad of J7.14
Figure 42. J4, J5 and J6 Shown on Map.15
Figure 43. J5 Orthophotoquad.15
Figure 44. J5 A Good OLS.15
Figure 45. J5 Area.15
Figure 46. J5 Another View.16
Figure 47. Map showing B4, B5, B6, JX3 Areas .16
Figure 48. B5 Similar to J4 .16
Figure 49. B5 Looking North.16
Figure 50. Map Showing OLSs B4, B5, and B6.16
Figure 51. OLS Software Image of B4 or JX3.16
Figure 52. Magnified Image of B4 or JX3.17
List of Figures (Continued)
Figure 53. B5 Software Output.17
Figure 54. B6 Software Output.17
Figure 55. Map of J4, J5, and J6 (Repeated from R 15).17
Figure 56. Map of Shiloh Valley Area, J4, J5, J6, JX3, B4, B5, B6, and B30.17
Figure 57. J4 Across the Creek from B5 .17
Figure 58. J4 Another View.18
Figure 59. J4 A Different Perspective.18
Figure 60. Image of J4.18
Figure 61. J4, J5, J7, and JX3 Orthophotoquad.18
Figure 62. J4 Software Output.18
Figure 63. J6 Software Output.18
Figure 64. Map showing B24, B32, B33, J13, and J16.19
Figure 65. Orthophotoquad of J13, J16, and J17.19
Figure 66. Orthophotoquad of J13 and J16.19
Figure 67. J16 (a), J13, J17, B24, B29, B32, and B33.19
Figure 68. J16 (b) Another View.20
Figure 69. J16 (c).20
Figure 70. J16 (d) From Other End.20
Figure 71. J16 (e) A Different View.20
Figure 72, J16 (f) Another View.20
Figure 73. Orthophotoquad of J17.20
Figure 74. J17 (a).21
Figure 75. J17 (b) A Different View.21
Figure 76. J17(c).21
Figure 11 .ill (d) Another View.21
Figure 78. J17(e).21
Figure 79. J17 (f).21
Figure 80. J17(g).22
Figure 81. Orthophotoquad Showing J13, J16, J17.22
Figure 82. B29.22
Figure 83. B29 Another View.22
Figure 84. B24 Software Output.22
Figure 85. B32 Software Output.22
Figure 86. B33 Software Output.23
Figure 87. J13 Software Output.23
Figure 88. J16 Software Output.23
Figure 89. J17 Software Output.23
Figure 90. Jll and J12 Shown on Map.24
Figure 91. Jll (a).24
Figure 92. J11 (b) A Different View.24
Figure 93 .Jll (c) Another View.24
Figure 94. Jll (d) A Different Perspective.24
Figure 95. Jll (e).24
Figure 96. J11 (f) A Different View.25
Figure 97. Jll (g).25
Figure 98. Jll Software Output.25
Figure 99. J12 Shown on Map.25
Figure 100. J12 Parallel to Road.25
Figure 101. J12 Paralleling Road.25
Figure 102. J12.26
Figure 103. Orthophotoquad of J12.26
Figure 104. J12 Software Output.26
List of Figures (Continued)
Figure 105. BIO Software Output.26
Figure 106. MapofB13 andB14.26
Figure 107. B14 (a) .26
Figure 108. B14 (b) Another View.27
Figure 109. B14 (c) A Different Perspective.27
Figure 110. B 14(d).27
Figure 111. B 14(e).27
Figure 112. B14(f).27
Figure 113. B14 (g).27
Figure 114. B13 Software Output.28
Figure 115. B14 Software Output.28
Figure 116. J2 Shown on Map.28
Figure 117. J2 Orthophotoquad.28
Figure 118. J2 (a) Rolling Terrain.28
Figure 119. J2 (b) A Different View.28
Figure 120. J2 (c).29
Figure 121. J2(d).29
Figure 122. J2 (e) Another View.29
Figure 123. J2 Software Output.29
Figure 124. Map of J14.29
Figure 125. J14.29
Figure 126. J14 (a).30
Figure 127. J14 (b) A Slightly Different View.30
Figure 128. J14 (c) Another view.30
Figure 129. J14 (d).30
Figure 130. J14 (e).30
Figure 131. B16 As Shown on Map .31
Figure 132. B16(a).31
Figure 133. B16 (b) Different Perspective.31
Figure 134. B16(c).31
Figure 135. B16(d).31
Figure 136. B16(e).31
Figure 137. B16 (f) Another View.32
Figure 138. B16(g).32
Figure 139. B16(h).32
Figure 140. B16.32
Figure 141. B37 Shown on Map.32
Figure 142. B37 (a) A Short Approach.32
Figure 143. B37 (b).33
Figure 144. B37 (c) .33
Figure 145. B37 (d).33
Figure 146. B37 Software Output.33
Figure 147. B36 Shown on Map.33
Figure 148. B36.33
Figure 149. B36 Another View.34
Figure 150. B36 Software Output.34
Figure 151. B35 Shown on Map.34
Figure 152. B35 (a).34
Figure 154. B35 (c) Another View.34
Figure 155. B35 Software Output.35
List of Tables
Table 1: OLS Sites
This work was funded by the United States Transportation Command (USTRANSCOM), managed by
the Air Mobility Command (AMC), and executed by the Control Systems Development and Applications
Branch of the Air Vehicles Directorate of the Air Force Research Laboratory (AFRL/RBCC).
Support for SynGenics was provided through the Simulation Technology Assessment (STA) Contract,
Prime Contract Number F33615-01-D-3105/0018, Delivery Order 18, under Subcontract Number D00058-
D6SC0578 to General Dynamics Advanced Information Systems (GDAIS).
The authors wish to express their appreciation to USTRANSCOM, AMC, AFRL, GDAIS, the Army's
Engineer Research and Development Center-Cold Regions Research and Engineering Laboratory
(ERDC-CRREL), and the Boeing Company for their participation in activities leading up to and
comprising this field demonstration.
Special thanks go to Dr. Charles E. Ryerson, ERDC-CRREL, for providing the photographs included
in this document.
1. Executive Summary
The objective of the Opportune Landing Site (OLS) Software Demonstration and Validation was to
enable and demonstrate the capability to locate possible suitable Landing Zones (LZs) that are smooth,
flat, firm, free of obstructions, and strong enough to support mobility aircraft operations. The Boeing
Company developed the OLS system software application to aid the warfighter in achieving global
access to the battlespace. The application currently comprises four separate modules of computer-coded
algorithms. One module uses satellite imagery to identify candidate landing areas that are large enough,
flat enough, and suitably free of vegetation, standing water, and obstacles. This module is referred to
in this report as the runway-finding software. Another module uses topographic data and historical
databases to determine soil type. A third module uses weather data and soil type to determine soil
moisture content, and the fourth module uses soil type and moisture content to determine soil strength.
A field demonstration and assessment of the OLS runway-finding software was held in St. Clair
County, IL, on 5 June 2007. The purpose of this portion of the OLS field demonstration was to assess the
capability of the runway-finding software. The field reviewing team comprised personnel from the Air
Force Research Laboratory (AFRL), the Air Mobility Command (AMC), the Army's Engineer Research
and Development Center-Cold Regions Research and Engineering Laboratory (ERDC-CRREL), General
Dynamics Advanced Information Systems (GDAIS), the Boeing Company, and SynGenics Corporation.
For the field demonstration, Boeing obtained LANDSAT imagery for St. Clair County, IL, collected
in May 2007. Boeing ran the OLS “flatness” software on the LANDSAT image to determine suitable
landing areas dimensioned 1,000 feet by 90 feet after the software was not able to find suitable landing
sites measuring 3,500 feet by 90 feet, according to the original requirement. Boeing then provided the
results of the software analysis of the region to the AMC. AMC designated a single trained Special
tactics team (STT) representative to identify all suitable landing areas dimensioned 3,500 feet by
90 feet in the same area (St. Clair County, IL) using aerial photography, topographic maps, digital
topographic elevation data (DTED), and other typically used means. Currently, the conventional "boots-
on-the-ground" method is used by STTs to review possible landing sites. The data used by the STT
representative was not of the same time frame as the satellite image, but was older by several years. On
5 June 2007, the field reviewing team visited many of the sites designated by the OLS runway-finding
software in order to assess the accuracy of the software and its capability to find suitable landing sites.
The team also visited 16 of the 17 sites proposed by the STT representative. The team went to 35 sites,
some of which contained clusters of candidate OLSs identified by either or both methods.
Of the 23 software-designated sites reviewed, all were considered potentially acceptable OLSs
although they were shorter in length than what was initially sought. Of the 17 STT-determined sites,
14 were considered potentially suitable. One of the sites appeared to cross power lines (although it
was difficult to determine from the dirt road to which the team had access). Two of the sites crossed a
construction site that was not reflected in the old DTED data. Additionally, one of those two also crossed
a ditch (on the opposite end from the construction site location.) Adjusting the software to require a
longer runway length may rule out some of the software-designated sites, but, based on the requirements
made for this demonstration, all of the software-designated sites proved acceptable. The implication is
that the runway-finding software module of the OLS System may provide an excellent tool in helping
the warfighter to achieve global access to the battlespace. Other modules of the system were not
Future steps may include further scientific investigation and refinement of this software module.
Additionally, the OLS Project Team continues to define potential uses for the OLS software and to
consider whether it should be distributed as a package or a service and who should maintain the database
of information upon which it relies, adding to and/or upgrading that database as situations change.
One issue is georegistration. That issue may provide a good candidate for future work under the OLS
Technology Maturation Plan.
The OLS Software Demonstration Plan describes a means to validate the utility and accuracy of the OLS
software application to locate and evaluate natural terrain LZs for airlift aircraft. The OLS application
uses satellite imagery to scan for obstacle-free, water-free and heavy-vegetation-free areas for evaluation
as candidate LZs. It then uses myriad data sources to infer soil type, and it uses mesoscale atmospheric
modeling and soil moisture modeling to infer soil strength. Areas that pass threshold values for openness;
absence of heavy vegetation, standing water, and obstacles; smoothness; and soil strength are identified
as opportune landing sites.
A proven OLS System will aid the warfighter in achieving anywhere-anytime access to the
battlespace. This technology will aid in conducting military operations from semi-prepared or
unprepared locations to effect a wide range of military options. Currently, these sites are evaluated
physically by military personnel before the planned operations begin. These evaluations may be
performed under hostile conditions. The OLS application was developed as an alternate method of site
evaluation. The OLS application will initially augment these physical site evaluations by prescreening
candidate areas, providing the benefits of reducing the initial search time, and limiting the number
of necessary physical site evaluations to the fewest areas. As technology and sensors improve, this
application is expected to eliminate the need for the physical evaluations.
A practical demonstration program highlighted the utility and accuracy of this module of the
application, with final results briefed to AMC in August 2007. The final report of the demonstration and
a Technology Maturation (Tech Mat) Plan were provided to AMC. The purpose of the demonstration
was to exercise the OLS software with respect to a set of criteria that represents a checkpoint along the
path toward a useful capability for airlift operations. This report covers a portion of the demonstration
program, describing the efforts on 5 June 2007.
2.1 Summary of Approach
The purpose of this portion of the OLS demonstration was to assess the capability of the runway-finding
software. The software was used to identify all suitable runways within an area of St. Clair County. In
addition, a manual inspection was performed using current conventional means, that is, identification of
sites by hand using satellite images and topographic maps. Sites were assigned numbers for identification.
Sites identified by inspection are designated with the prefix "J M , while those determined by the software
are named with a "B" prefix. There is no relationship implied between sites having the same numbers
but different prefixes. On 5 June 2007, the observation team drove to most sites identified and visually
inspected/verified their suitability as a landing zone. Each stop was identified by its quadrant number.
This report details the results of this portion of the demonstration and compares the findings of the
observation team with respect to each candidate OLS visited.
The 5 June demonstration was intended to showcase the capabilities of the OLS software to the
AMC staff, demonstrate the current state of the technology, and reveal the potential of the technology
for further development and fielding. Further objectives were to prove that Key Performance Parameters
(KPPs) and exit criteria for the OLS software demonstration and validation program have been met and
to lay the foundation for the technology maturation and risk-mitigation way forward. The purpose of this
portion of the OLS demonstration was to assess the capability of the runway-finding software
2.2 Identifying Candidate OLSs
The OLS algorithms make some specific assumptions about the physics of reflected electromagnetic
radiation to find suitable landing sites. Appreciating these assumptions is important in understanding the
capabilities and limitations of the application. Multispectral and hyperspectral satellite imagers measure
the electromagnetic radiation emitted from the sun and reflected by a given area (pixel) of the earth’s
surface. The reflected component also includes atmospheric scattering of solar radiation, and, as the
spectra approaches the IR spectral region, the radiation at the sensor includes earth- and atmospheric-
emitted radiation. This radiation is formatted by the imager into separate images based on the wavelength
of the radiation.
The OLS algorithms are based on the assumption that variations of the earth’s surface reflectance are
caused by physical (spatial) and material (spectral) characteristics, which can be used to discriminate
the spatial and spectral properties of the terrain for a given area (pixel). These variations are used
to identify standing water, areas containing heavy vegetation (high chlorophyll), and uneven terrain
(combined spatial/spectral inhomogeneity). Conversely, areas with highly uniform reflectance (spatial
and spectral homogeneity), are assumed to be flat areas of like material substance (dirt, grass, rock, etc.).
The algorithms reject areas with large variations in reflectance, such as those caused by sharp contrast
between the asphalt of a road or runway and the surrounding soil or vegetation; the OLS application
looks solely for areas of homogeneous natural terrain. This report details the results of this portion of the
demonstration and compares the findings of the observation team with respect to each candidate OLS
The red rectangle on the map below indicates that portion of St. Clair County in which the
demonstration took place.
i rear t • p&nwnr mm 3 i-hjj;
Figure 1. Demonstration Area within St. Clair County
3. Methods, Assumptions, and Procedures
It was agreed that the objective of the OLS Software Demonstration and Validation Program would be
shown to have been met if the team were to demonstrate that the documented exit criteria were met. The
purpose of this portion of the demonstration was to assess the OLS Software against two performance
criteria, one of which was an exit criterion and a KPP:
• KPP P01: Capability to identify suitable landing sites in a specified area, given that suitable landing
sites exist. Suitable is defined as having an area of the specified dimensions that is flat and free of
obstacles, standing water, and heavy vegetation. Bearing strength is not a consideration for suitability
in this context. Exit criterion: at least 50 percent of OLSs found. Objective: 100 percent.
• P03: Low incidence of false positives. Probability of designating an unsuitable landing site as a suitable
OLS—a measure of accuracy expressed as the percentage of OLSs identified by the software that were
unsuitable. Suitability as defined for this criterion excludes bearing strength. The value with respect
to this desirement was to be assessed through comparison of the software analysis results with field
inspection and observation results for St. Clair County. The goal was 0 percent. No upper bound was
set at this stage.
Results are highlighted in Section 4. Results. KPPs P01 and P03 comprised the focus of the 5 June
effort. Boeing obtained LANDSAT imagery for St. Clair County collected in May 2007. Boeing ran the
OLS “flatness” software on the LANDSAT image and determined suitable landing areas dimensioned
1,000 feet by 90 feet. Boeing used those dimensions with AMC approval, after reporting that the software
did not find any suitable landing sites 3,500 feet in length. Boeing then provided the results of the
software analysis of the region to AMC.
In parallel with the software analysis, AMC tasked a representative of a STT, to identify all suitable
landing areas dimensioned 3,500 feet by 90 feet in St. Clair County using aerial photography of the same
area as the LANDSAT imagery, topographic maps, DTED, and other means typically used by STTs. This
method is henceforth referred to as ’’inspection”. The data used by the STT representative was not of the
same time frame as the satellite image, but was older by several years. For example, Mid America Airport
was under construction in the STT data, yet was operational by the time the test was conducted. The STT
found OLSs by inspection in only eastern St. Clair County, looked for OLSs measuring 3,500 feet by 90
feet, and found some longer ones as well.
An AMC-designated referee was tasked to compare the software results with those of the STT
representative using inspection and to calculate the percentage of correct sites (P01) and the incidence
of false positives (P03). Other participants in the demonstration included representatives of AMC, AFRL/
RBCC, GDAIS, SynGenics, ERDC-CRREL, and the Boeing Project Manager. They confirmed sites
by observation. AFRL/RBCC representatives along with ERDC-CRREL served as impartial obervers
and adjunct referees. SynGenics served as observer and recorder, and the AMC representative was
the photographer. As required by the Demonstration (demo) Plan, the team obtained vantage points
as close as possible to the location of the alleged OLS and ascertained by observation whether the site
was a suitable LZ. The percentage of suitable LZs was to be recalculated based on these findings. LZs
identified by the software, missed by inspection, but subsequently confirmed by observation would
contribute to both the numerator and the denominator of this calculation. The software performance was
to be considered successful if it found at least 50 percent of the suitable sites.
Although not a KPP for the demonstration, a low incidence of false positives was desired. A false
positive occurs when the software designates an area as a suitable OLS when, in fact, it is unsuitable. The
occurrence of false positives would be calculated based on the comparison of software-identified LZs
with inspection and observation results.
4. Results and Discussion
The demonstration team followed the Demonstration Plan as detailed in the Demo Plan section 5.1 except
that they visited nearly every site found by either the software or the AMC-trained individual, the STT
representative. Exception: The team visited only eastern St. Clair County.
• KPP P01: Capability to identify suitable landing sites proved difficult to quantify because it was
unknown how many suitable sites exist in the region chosen for the demonstration of this desirement.
OLS-MS identified 40 sites, whereas an individual using the standard manual method identified only
17 sites in the region. It could be argued that the software scored 235 percent. While the exact score is
unknown, there is agreement that the exit criterion of at least 50 percent was certainly exceeded, and it
could be argued that the objective of 100 percent was met. The lesson learned is that properly defining
the measurand and the method of collecting the data to support quantification against that measurand
• P03: Incidence of false positives was 0, meeting the objective.
The following pictures depict sites visited. They are listed in the order visited. Table 1, OLS Sites,
summarizes the visits. Information in each header includes the site designation; coordinates at the
northwest corner of the landing zone; runway magnetic heading (degrees); and length (feet). Pictures
comprise 1) the National Geographic map, 2) photo(s) of the field, 3) orthophotoquad (for J numbers), 4)
OLS software output. Text reflects findings of the observation team concerning the site. The runway¬
finding software identified 40 candidate LZs. Some included clusters of possible runways 1,000 feet or
longer, for a total of 54 potential OLSs. The inspection method identified 17 possible runways that were
at least 3,500 feet long. Of the 16 STT-determined sites visited, 13 were considered potentially suitable.
One of the sites appeared to cross power lines (although it was difficult to determine from the dirt road to
which the inspection team had access). Two of the sites crossed a construction site that was not reflected
in the old DTED data. Additionally, one of those two also crossed a ditch on the opposite end from the
construction site location. The combined results of the evaluation for both capability to identify landing
sites and the occurrence of false positives indicated that the runway-finding software performed very well
in the portions of the demonstration that have been completed and, in combination with other tools, could
provide an excellent means of finding potential OLSs.
Table 1 indicates sites that the team visited and documented, in the order visited.
Table 1: OLS Sites
Bl: 38° 38' 04.53" N 89° 47' 01.88" W, 180|360
B2: 38° 37' 8.31" N 89° 41' 19.86" W, 180|360
B3: 38° 33' 31.54" N 89° 42' 34.80" W, 180|360,
J15: 38° 39' 9.6” N 89° 45' 16.4" W, 100|280, 4600
B28: 38° 39' 18.65" N 89° 46' 54.82" W, 90|270
JX1: 38° 33' 43.4" N 89° 48' 24.2" W, 100|280,
J8: 38° 33' 35.93" N 89° 49' 5.34" W, 120|300,
JX2: 38° 33' 35.5" N 89° 49' 4.5" W, 110|290, 3003
Table 1: OLS Sites
J7: 38° 31 ? 34.7” N 89° 52' 55.4” W, 150|330, 3000
J5: 38° 30' 09” N 89° 52' 47” W, 180|360, 3100 ft.
JX3: 38° 31' 10.49” N 89° 51' 24.57” W
J4: 38° 30' 39.1” N 89° 53' 29.1” W, 90|270, 3600
J5: 38° 30' 09” N 89° 52' 47” W, 180|360, 3100 ft.
J6: 38° 35' 29.57” N 89° 29' 19” W, 80|260, 4000
B5: 38° 30' 34.55” N 89° 51' 21.91” W, 180|360
B4: 38° 31' 10.49” N 89° 51' 24.57” W, 90|270, 1000
B6: 38° 30' 04.64 N 89° 51' 51.68” W, 180|360
J13: 38° 29' 39.3” N 89° 42' 44.3” W, 174°, 4200 ft.
J16: 38° 29' 16.4” N 89° 43' 40.4” W, 47°, 3500 ft.
J17: 38° 28' 46.6” N 89° 43' 23.1” W, 144°, 3800 ft.
B24: 38° 25' 16.47” N 89° 47' 15.59” W, 060|240
B29: 38° 31' 26.06” N 89° 47' 22.24” W, 090|270
B32: 38° 30' 32.13” N 89° 43' 32.47” W, 090|270
B33: 38° 30' 23.38” N 89° 43' 32.14” W, 0901270
Jll: 38° 33' 48.4” N 89° 48' 24.2” W, 100|280,
J12: 38°32'29.3” N 89°47'5.4” W, 196°, 3100 ft.
BIO: 38° 32' 34.64” N 89° 46' 21.88” W, 180 360
B13: 38° 22' 52.29” N 89° 48' 3.21” W, 360|180
B14: 38° 22' 54.14” N 89° 46' 4.61” W, 360|180
J2: 38° 16' 43.1” N 89° 45' 57” W, 180|360, 3800
J14: 38°16'42.8” N 89°42'32.3” W, 109°, 3800 ft.
B16: 38° 17' 34.78” N 89° 43' 26.76” W, 180|360
B37: 38°20'49.33” N 89°48'18.24” W, 90|270
B36: 38° 23' 16.01” N 89° 54' 27.38” W, 90|270
B35: 38°24'37.58” N 89°48'35.72” W, 90|270.
Visit 1—Bl: 38° 38' 04.53" N 89° 47' 01.88" W, 180|360
B1 was deemed a good OLS.
Figure 2. Bl Shown on Map
Figure 3. Bl
Figure 5. B1 Software Output
Visit 2—B2: 38° 37 f 8.31" N 89° 41 ' 19.86" W, 180|360
B2 was very good. It could be oriented 90|270 or 180|360 degrees. A good approach was noted. The STT
representative said he did not find it because it was an east-west runway that extended beyond the eastern
edge of the area of consideration.
Figure 10. B2 (d) Acceptable Approach
Figure 11. B2 (e)
Figure 13. B2 Software Output
Visit 3—B3: 38° 33' 31.54" N 89° 42' 34.80" W, 180|360
B3 was also a good OLS.
Figure 14. B3 Shown on Map
Figure 15. B3
Figure 16. B3 Another View Figure 17. Orthophotoquad of B3
Figure 18. B3. Software Output
Visit 4—J15: 38° 39' 9.6" N 89° 45' 16.4" W, 100|280, 4673 ft.
B28: 38° 39’ 18.65" N 89° 46' 54.82" W, 90|270
These runways are both oriented roughly east-west and are equivalent in terms of landing suitability. J15
and B28 are located approximately one field apart.
Figure 19. Bl, and J15 As Shown on Map
Note: See red 15
Figure 20. J15
Figure 21. Orthophotoquad of J15
Figure 22. J15 Software Output
Visit 5—JX1: 38° 33’ 43.4” N 89° 48' 24.2" W 110|280, 3419 ft.
JX1 is parallel to JX2 and J8. See Visit 6.
Figure 23. JX1 Shown on Map.
Note: Fingers Point to the OLS
Figure 24. JX1
Figure 25. JX1 From Another Direction
Figure 26. JX1 Another View
Visit 6—J8: 38° 33 ? 35.93" N 89° 49 f 5.34" W, 120|300, 3118 ft.
JX2: 38° 33 f 35.5" N 89° 49 f 4.5" W, 110|290, 3003 ft.
These areas are under construction. There is a ramp from 1-64 into what will be Hayden Retail Office
Park. The construction began after the image was taken to identify the site; so it is reasonable that the
STT would not have ruled the site out because of construction. However, JX2 crosses a ditch, which
makes the site unacceptable (see Figure 30), both because of the ditch and because of working with
old data which did not show the construction site. The STT affirmed that he took a chance on this one,
thinking the ditch might be a road. A higher resolution image would have revealed the truth, ’’Which is
why you put boots on the ground,” the SST commented. This software output shows the OLS crossing
the runway at MidAmerica Airport, whereas Figures 32. 34, and 36 show that it does not, illustrating the
georegistration problem. Figure 32 shows that the OLS approaches the airport runway.
Figure 29. J8 A Closer View
Figure 31. JX2 (a)
Figure 32. JX2 (b)
Figure 33. JX2 (c)
Figure 34. JX2 Repeat of Orthophotoquad
Figure 35. JX2 (d)
Figure 36. J8
Note: Software Output Falsely Indicating OLS Runway
Crossing at MidAmerica Airport, an Illustration of Geo¬
Visit 1 — 31 : 38° 31 ? 34.7" N 89° 52 ? 55.4" W, 150|330, 3000 ft.
The site has a ditch, but the OLS runs east of the ditch. A house is situated the corner of the LZ area, but
neither the ditch nor the house renders the area unacceptable.
Figure 38. J7
Note: Picturing OLS Running East of the Ditch and There¬
Figure 39. J7 Another View
Figure 40. J7 A Different Perspective
Figure 41. Orthophotoquad of J7
Note: The site has a house at the corner of the LZ area, but
still the area is acceptable
Visit 8—J5: 38° 30 ? 09" N 89° 52 f 47" W, 180|360, 3100 ft.
B5: 38° 30 f 34.55" N 89° 51 f 21.91" W, 180|360
B4: 38° 31 T 10.49" N 89° 51 T 24.57" W, 90|270,1000 ft.
B6: 38° 30 f 04.64 N 89° 51 f 51.68" W, 180|360
JX3: 38° 31 f 10.49" N 89° 51 f 24.57" W, 180|360
J4: 38° 30 ? 39.1" N 89° 53 f 29.1" W, 90|270, 3600 ft.
J5: 38° 30 ? 09" N 89° 52 f 47" W, 180|360, 3100 ft.
J6: 38° 35 f 29.57" N 89° 29 ? 19" W, 80|260, 4000 ft.
Sites B4, B5, B6, JX3, and J4-J6 were all very close together. The software-identified sites might have
been missed by the STT representative because he was looking only for areas at least 3,000 ft. long. J4
is one of the OLSs the STT representative found on his way to work and was not among the OLSs that
he found in his workbook. The STT representative says the best way to land is on a 180° heading. B6 is
only 1,000 to 2,000 feet. The runway-finding software did not like the creek or dip at the south end. The
Boeing PM pointed out the dip at the far end of the runway and questioned whether the dip is why the
software rejected this as a candidate LZ when it was looking for only 1,000 feet.
J4 is not precisely the same as B5. It is in the same field but on the other side of the creek. The
team did not find OLSs west of these sites because of the boundary of the search space. J7 has a ditch,
but the OLS runs east of the ditch. It is nearly the same as B4. One team member observed that J7 is
dangerously close to the tree line, which is on the other side of the road and not very visible in either
image. The photos of B4, B6, and JX3 washed out due to a camera malfunction; hence, there is no
documentation of B4-B6. All were good sites. B4 and J7 align, B5 and J4 align, and B5 and J6 align.
Figure 48. B5 Similar to J4
Note: Different Part of Same Field
Figure 47. Map showing B4, B5, B6, JX3 Areas
Note: Focus Is on B5, Section 4
Figure 49. B5 Looking North
Figure 50. Map Showing OLSs B4, B5, and B6
B | p _ | g i lisp *2 Hats: 0 -V
Figure 51. OLS Software Image of B4 or JX3.
Note: Software Utility Indicates a Cluster.
Figure 52. Magnified Image of B4 or
Figure 54. B6 Software Output
Figure 56. Map of Shiloh Valley Area, J4, J5, J6, JX3,
B4, B5, B6, and B30
Figure 53. B5 Software Output
Figure 55. Map of J4, J5, and J6 (Repeated from P. 15)
Note: J7 Indicated by Arrow
Figure 57. J4 Across the Creek from B5
Figure 60. Image of J4
Figure 61. J4, J5, J7, and JX3 Orthophotoquad
Figure 62. J4 Software Output
Figure 63. J6 Software Output
Visit 9—B24: 38° 25’ 16.47" N 89° 47' 15.59" W, 060|240
B29: 38° 31' 26.06" N 89° 47' 22.24" W, 090|270
B32: 38° 30' 32.13" N 89° 43' 32.47" W, 090|270
B33: 38° 30' 23.38" N 89° 43' 32.14" W, 090|270
J13: 38° 29' 39.3" N 89° 42' 44.3" W, 174°, 4200 ft.
J16: 38° 29’ 16.4" N 89° 43’ 40.4" W, 47°, 3500 ft.
J17: 38° 28' 46.6" N 89° 43’ 23.1" W, 144°, 3800 ft.
Areas B24, B29, B32, B33, J13, J16, and J17 are quite near each other and so are grouped here. B17, J13,
J16 and J17 are near Miscoutah.
Figure 66. Orthophotoquad of J13 and J16
Figure 67. J16 (a), J13, J17, B24, B29, B32, and B33
Note: Sites are close to each other
Figure 74. J17 (a)
Figure 75. J17 (b) A Different View
Figure 76. J17 (c)
Figure 77. J17 (d) Another View
Figure 78. J17 (e) Figure 79. J17 (f)
Figure 80. J17 (g)
Figure 82. B29
Figure 84. B24 Software Output
Figure 81. Orthophotoquad Showing J13, J16, J17
Figure 85. B32 Software Output
Figure 87. J13 Software Output
Figure 86. B33 Software Output
Figure 88. J16 Software Output
Figure 89. J17 Software Output
Visit 10—Jll: 38° 33 ? 48.4" N 89° 48 f 24.2" W, 100% 3500 ft.
J12: 38°32'29.3" N 89°47'5.4" W, 196% 3100 ft.
B10: 38° 32' 34.64" N 89° 46' 21.88" W, 180|360
Jll and J12 are near B10 along with J8. JX1 and JX2 are parallel to J8. Jll parallels the road on which
the team stood. The church was photographed in the wrong direction at first. The team was initially
concerned that the possibility existed that the OLS was crossed by power lines. They could not confirm
from the roadway, but ultimately deemed the site acceptable based on the orientation of the OLS. The
SynGenics representative asked why the software did not find the field photographed. It turned out that
there was a six-inch high patch of vegetation in part of the area. J12 is north-south, parallel to the road,
and is a good site.
Figure 90. Jll and J12 Shown on Map
Figure 92. Jll (b) A Different View
Figure 91. Jll (a)
Note: Shown Here Near BIO, J8, and J12
Figure 93. Jll (c) Another View
Figure 94. Jll (d) A Different Perspective Figure 95. Jll (e)
Figure 96. Jll (f) A Different View
Figure 97. Jll (g)
Figure 98. Jll Software Output
Figure 100. J12 Parallel to Road.
Figure 99. J12 Shown on Map
Figure 102. J12
Figure 104. J12 Software Output
Figure 103. Orthophotoquad of J12
Figure 105. BIO Software Output
Visit 11—B13: 38° IT 52.29" N 89° 48 ? 3.21" W, 180|360
B14: 38° 22 ? 54.14" N 89° 46 ? 4.61" W, 180|360
B14 was very good. The STT representative did not find OLSs to match them. There was high corn on the
site at the time of the visit; so the team could not see the field well.
Figure 108. B14 (b) Another View
Figure 109. B14 (c) A Different Perspective
Figure 110. B14 (d)
Figure 111. B14 (e)
Figure 112. B14 (f)
Figure 113. B14 (g)
Figure 114. B13 Software Output
Figure 115. B14 Software Output
Visit 12—J2: 38° 16 ? 43.1" N 89° 45 ? 57" W, 180, 3800 ft.
The Boeing software did not find this one. This was the third failure of the STT-designated landing sites.
Power lines and telephone lines as well as rolling terrain and vegetation are reasons it was not considered
viable. The Boeing software did not designate it for these reasons.
Figure 116. J2 Shown on Map
Figure 117. J2 Orthophotoquad
Figure 118. J2 (a) Rolling Terrain
Figure 119. J2 (b) A Different View
Figure 122. J2 (e) Another View Figure 123. J2 Software Output
Visit 13—J14: 38°16 ? 42.8" N 89°42 ? 32.3" W, 109°, 3800 ft.
Lehr Road has a suitable Landing Zone. The Software ruled it out because of vegetation (a winter wheat
crop in April). A solution might include (1) reducing the vegetation threshold or (2) georectification (-200
Figure 124. Map of J14
Figure 125. J14
Note: Software Illustrates Georegistration Problem
Figure 126. J14 (a)
Figure 127. J14 (b) A Slightly Different View
Figure 128. J14 (c) Another view
Figure 129. J14 (d)
Figure 130. J14 (e)
Visit 14—B16: 38° IT 34.78" N 89° 43 ? 26.76" W, 180|360
B16 is a good Landing Zone space, and it is very wide. There is a cluster of suitable landing zones five
pixels in width, indicating numerous potential sites.
Figure 131. B16 As Shown on Map Figure 132. B16 (a)
Figure 137. B16 (f) Another View
Figure 139. B16 (h)
Figure 138. B16 (g)
Figure 140. B16
Note: Arrow Indicates Cluster of LZs in Software Output
Visit 15—B37: 38°20 ? 49.33" N 89°48 ? 18.24" W, 90|270
B 37 is in the Kaskaskia River flood plain. The team scored this as a good site based on what the software
was designed to do. The site avoided power lines because it was beside the road. However, per the referee,
the site is surrounded by trees. It was suggested that future versions of the software should consider
approach and departure space, in which case this site might no longer be considered suitable.
Figure 141. B37 Shown on Map
Figure 142. B37 (a) A Short Approach
Figure 143. B37 (b)
Figure 145. B37 (d)
Figure 144. B37 (c)
Note: Arrow Indicates Fence in Lower Right Corner.
Figure 146. B37 Software Output
Visit 16—B36: 38° 23 ? 16.01" N 89° 54 ? 27.38" W, 90|270°
B36 is an east-west site. The team could not get really close, but it looks good. The team photographed
one OLS that starts near the big barn and one that starts beyond the cornfield.
Figure 147. B36 Shown on Map Figure 148. B36
Figure 149. B36 Another View
Figure 150. B36 Software Output
Visit 17—B35: 38°24 f 37.58" N 89°48 ? 35.72" W, 90|270°.
B35 was a very good site. There were also others on the other side. While searching for B35, the team
went past a flooded strip mining area which the software did not identify as OLSs.
Figure 152. B35 (a)
Figure 151. B35 Shown on Map
Figure 153. B35 (b)
Figure 154. B35 (c) Another View
The inspection team members viewed 40 sites during the day on 5 June 2007. In general, the software was
judged to have performed quite well. Of the 23 "B" sites reviewed, all were considered potentially acceptable
OLSs, although they were shorter in length than what was initially sought. The Inspection method identified
17 possible runways that were at least 3500 feet long, of which three were rejected. Two had become
construction sites in the period between the taking of the satellite images and the inspection. Another was
considered unsuitable because of the presence of power lines and a ditch. Adjusting the software in order
to require a longer runway length might have ruled out some of the "B" sites, but, in accordance with the
requirements for this demonstration, all of the software-designated sites proved acceptable. Based on this
sampling and its results, it would appear that the OLS runway-finding software could be used to identify
potential OLSs at least as well as the methods currently being used by STTs, given that the software database
is properly maintained.
It was agreed that the objective of the OLS Software Demonstration and Validation Program would
be shown to have been met if the team were to demonstrate that the documented exit criteria were met.
In terms of KPP P01: Capability to identify suitable landing sites in a specified area, given that suitable
landing sites exist, the runway-finding software identified 40 sites, whereas an individual using the
standard manual method identified only 17 sites in the region. It could be argued that the software scored
235 percent. While the exact score is unknown, there is agreement that the exit criterion of at least 50
percent was certainly exceeded, and it could be argued that the objective of 100 percent was met. The
lesson learned is that properly defining the measurand and the method of collecting the data to support
quantification against that measurand is important. P03, Incidence of false positives was 0, meeting the
objective. This result exceeded expectations.
The performance of the software vastly exceeded the exit criteria. Moreover, it met the stretch goals
established in the Demonstration Plan.
Future steps may include further scientific investigation and refinement of the software module. The
OLS Project Team continues to explore potential additional uses for the OLS software and its capabilities.
Other issues to be resolved include whether it should be distributed as a package or a service, and who
should maintain the database upon which it relies, adding to and/or upgrading that database as situations
change. Resolving the issue of georegistration is a necessary step for possible future development under
the OLS Technology Maturation Plan, AFRL-RB-WP-TR-2008-3064 (AD number B336859). A more
complete view of recommendations to move toward an OLS Initial Operational Capability can be found
in that document.
LIST OF ACRONYMS, ABBREVIATIONS, AND SYMBOLS
Air Force Base
Air Force Research Laboratory
Air Vehicles Directorate of AFRL
Control Sciences Dibision of AFRL/RB
Control Systems Development & Application Branch of AFRL/RBC
Air Mobility Command
Digital Topographic Elevation Data
Engineer Research and Development Center-Cold Regions Research and
Development Center (Army)
General Dynamics Advanced Information Systems
Infrared (spectral region)
Key Performance Parameter
Land Remote Sensing Satellite System
Opportune Landing Site
Special Tactics Team
Technology Maturation (Plan)