See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/224119569

Unmanned marine vehicle use at Hurricanes Wilma and Ike

Conference Paper · November 2009

DOI: 10.23919/OCEANS.2009.5422201 · Source: IEEE Xplore

CITATIONS READS
37 790

4 authors, including:

Eric Steimle Robin R. Murphy

20 PUBLICATIONS 806 CITATIONS

Texas A&M University
302 PUBLICATIONS 11,818 CITATIONS
SEE PROFILE
SEE PROFILE

Michael Lindemuth
University of South Florida
12 PUBLICATIONS 389 CITATIONS

SEE PROFILE

All content following this page was uploaded by Eric Steimle on 09 June 2014.

The user has requested enhancement of the downloaded file.

 Unmanned Marine Vehicle Use at Hurricanes Wilma
and Ike
Eric T. Steimle
Research Professor, ESP&G,
University of South Florida
140 7th Ave S. St. Petersburg, FL 33701.

Robin R. Murphy
Raytheon Professor, Computer Science and Engineering
Texas A&M,
College Station, TX 77843

Michael Lindemuth
Software Engineer, Center for Ocean Technology
University of South Florida
140 7th Ave S. St. Petersburg, FL 33701

Micheal L. Hall
Project Engineer,
Applied Environmental and Ocean Sciences, LLC.
2534 22nd Street N, St. Petersburg, FL 33713

Abstract- The Center for Robot-Assisted Search and Rescue has fog and spray interfere with autonomous sensing and avoidance
fielded two AEOS man-portable unmanned surface, a YSI as well as teleoperation.
Ecomapper and VideoRay underwater marine vehicle in the
aftermath of Hurricanes Wilma (2005, Florida) and Ike (2008, 1. INTRODUCTION
Texas) for littoral structural inspection. By providing above and
below waterline inspection of roads, bridges, seawalls, robots While hurricanes are associated with large scale
could cut the time involved in determining how to repair a bridge
search and rescue activities on land, inspection of coastal
from on order of a month to a few days and could eliminate the
need for divers to risk their lives trying to see and move around littoral regions is also important. Bridges must be inspected, as
debris in cloudy, fast-moving currents. For example, using a they are needed for responders and recovery workers to have
DIDSON acoustic camera, healthy pilings were seen at Rollover access to affected areas, and because they influence the
Pass Bridge but scour was seen at the Marco Island bridge. general recovery of the area. Seawalls, levees, or dikes may be
The AEOS unmanned surface vehicle is a catamaran compromised and create a secondary disaster such as seen in
style design with a central T-shaped chassis that supports the New Orleans at Hurricane Katrina. Channels must be restored
electronics, instrumentation and control system. The footprint is and docks repaired as part of the economic recovery and
1.9m long and 1.2m wide with a minimum draft of 0.3m. The restoration of shipping.
USV is engineered for autonomous and teleoperated control. The
This paper describes the use of surface and
custom software allows for point and click waypoint inputs
through a Google Earth style interface. All data and system underwater unmanned marine vehicles (UMV) for post-
information is georeferenced and databased. disaster bridge inspection by the Center for Robot-Assisted
Besides illustrating the utility of unmanned marine Search and Rescue (CRASAR) in the aftermath of Hurricane
vehicles for hurricane damage assessment, the field work Ike. CRASAR is a center at Texas A&M that promotes robots
identified many open research questions for unmanned surface for emergency response and sends out volunteer teams of
vehicles (USVs). Fully autonomous control in USVs is hampered scientists and robot manufacturers to disasters. Hurricane Ike
by the loss of GPS and inaccuracies near structures, such as was a Category 4 storm that struck Galveston, Texas on
going under bridges as well as rough seas and currents. Even September 13, 2008. The Rollover Pass Bridge on the adjacent
with accurate GPS, the simultaneous localization and mapping
Bolivar peninsula, a major artery to the area, was severely
(SLAM) for super- and sub-structures is difficult. USVs suffer
from the limitations of wireless communications, including damaged. In December, 2008, the Texas Transportation
intermittent loss and low bandwidth compared with the amount Institute (TTI) and the Texas Department of Transportation
of video imagery being transmitted. Atmospheric effects such as permitted CRASAR to test UMVs for bridge inspection. In
particular, the agencies wanted an evaluation of how well the

0-933957-38-1 ©2009 MTS

Sea-RAI unmanned surface vehicle (USV) could inspect the World Trade Center disaster under the invitation of the State
bridge footings for scour (erosion and separation from the of New York18 and with Florida Task Force 3 at Hurricane
bottom) and map the debris field around the bridge (i.e., is Katrina. CRASAR deployed Micro Air Vehicle for structural
there debris likely to be pushed into the substructure and inspection of damage from Katrina to multistory commercial
damage it). The Sea-RAI at the Rollover Pass Bridge is shown structures along the Mississippi Gulf Coast19
in Figure 1. Unmanned marine vehicles have not been widely
As described in,1 the inspection of the underwater used for disaster response,3 though they have been for other
portion of a bridge, called the substructure, is currently done applications, including mapping the Great Barrier Reef,4
manually with divers who must work in high currents, low demining,5 inspection of cables,6 sampling sea-air interfaces7,
visibility, and debris to physically see and touch damage. sediments8 and for high resolution mapping of estuaries, lakes
Manual inspection puts the divers at risk and is not efficient, and rivers. With the notable exception of marina mapping,9
for example due to the tidal currents at the Rollover Pass these littoral operations do not take place in close proximity to
structures or constricted areas. The deployment at Rollover
Pass Bridge is an extension of previous field work and
research by CRASAR. CRASAR conducted a survey of the
Marco Island, Florida, bridge in the wake of Hurricane
Wilma2 with an AEOS man-portable surface vehicle. That
work led to the Sea-RAI project which refined the AEOS
platform, added navigational autonomy, and improved human-
robot interfaces.10 As far back as 1991, UMVs have been
proposed for post-disaster inspection11 however, the post-
Hurricane Wilma deployment is the first known application of
a USV for disaster response.

III. DEPLOYMENT

The six person CRASAR team was in the field from

Dec. 17-19, 2008, and was comprised of roboticists, civil
engineers, and responders and three unmanned marine
vehicles. The six person team consisted of two roboticists
Figure 1 Sea-RAI near Rollover Pass Bridge (Murphy, Steimle), both of whom had participated in the
Hurricane Wilma deployment, two civil engineering
bridge, divers could only work for 15 minutes at the change of professors (Hurlebaus, Medina-Cetina), a responder from the
each tide. CRASAR showed the efficacy of unmanned surface Texas Engineering Extension Service (May), which is the
vehicles (USV) in 2005 at Hurricane Wilma2 for determining state agency for urban search and rescue, and a graduate
scour and locating debris, providing a foundation for this student (Lindemuth). In addition, YSI sent two experts (Hall,
deployment. Hurricane Wilma, a category 5 storm, made Slocum) to use the Ecomapper UUV with CRASAR on Dec.
landfall at Cape Romano, Florida on October 25th 2005. 18. The primary vehicle was the Sea-RAI unmanned surface
CRASAR was invited by the Marco Island Yacht Club (MIYC) vehicle, with a VideoRay ROV and YSI UUV Ecomapper as
on Marco Island to investigate damage to their building, docks secondary vehicles for experimentation and testing. The team
and seawall. focused on the USV because as noted in2 surface vehicles
This paper reviews the related work on UMVs for have important advantages over underwater vehicles: they can
littoral inspection, and then describes the CRASAR be more accurately controlled and localized through GPS, they
deployment. The deployment was successful both in terms of can carry a larger payload, and they can continuously
meeting the two mission objectives but also in identifying broadcast data to observers in real-time.
challenges in vehicle control, multi-robot coordination, The deployment had a primary mission for the
human-robot interaction, and sensing. It contributes an agencies and a secondary scientific mission. The primary
understanding of the post-disaster bridge inspection which can mission was to evaluate the utility and performance of the
be applied to damage from natural events (e.g., hurricanes and USV for two tasks: inspection of the bridge substructure and
flooding) but also to man-made incidents such as the 2007 I- mapping of the debris field. The Sea-RAI was able to meet
35W bridge collapse in Minneapolis. both objectives. The primary mission success was the
acquisition of comprehensible underwater imagery deemed of
II. RELATED WORK use to the structural community. The secondary scientific
mission consisted of a preliminary Viewpoint-Oriented
The use of unmanned systems for rescue and disaster Cognitive Work Analysis (CWA)12 to determine how a USV
response is still experimental, and the work presented here is might be used in the future and experimentation with the two
novel. Ground robots were deployed by CRASAR at the 2001 UUVs. CWA provides a systems perspective of a work
activity, in this case how UMVs would actually be used. The adaptation of the AEOS platform built for environmental
CWA was the basis for 1) an evaluation of technology mapping. The Sea-RAI carries a DIDSON acoustic camera for
transition potential and roadmap based on robotics, response, subsurface inspection and three video cameras (forward, rear,
and civil engineering expertise and 2) an assessment of hemispherical) for viewing above the waterline. The robot can
resilience in human-robot interaction. carry additional sensors17 but for this specific mission only the
DIDSON was outfitted. The Sea-RAI is battery powered and
IV. ROLLOVER PASS can operate for 4-6 hours depending on the currents and
payload. A notable feature of the Sea Ray is that it stores
Rollover Pass is a man-made straight that cuts sensor data and the internal state of the robot, creating a
through the Bolivar Peninsula and links the Gulf of Mexico database that can be displayed using standard mapping
with Rollover Bay and East Bay on the upper Texas coast in techniques. Mike Lindemuth has developed a Google Earth
eastern Galveston County. The Rollover Pass Bridge is a two- interface shown in Figure 3, which displays the information in
lane concrete span on Texas Highway 87 connecting a 200 ft an easy to interpret manner. Data at any point in time or
channel between the Gulf of Mexico, Galveston Bay, and the location can be retrieved with a click. The VideoRay ROV
intercoastal waterway. The pass is subject to intense tides and and YSI Ecomapper were also used. The Videoray is already
turbidity. The tides can reach an excess of 6 knots with zero in use by some departments of transportation for visual
visibility in the water. The slack tide lasts for less than 15 inspection of bridges and debris, either lowered off a bridge or
minutes, with a current under two knots for an hour on both from a boat. The YSI Ecomapper carries a side-scan sonar
sides of the slack. The Bridge has a weir built on the suitable for mapping debris fields.
Galveston Bay side of the pass. This can be seen in Figure 2.
Due to the large of amount of water that must pass though this
channel, the difference in water level on either side of the
Weir can reach ~20 cm.

Figure 3 Google Earth display of deployment data

VI. MISSION RESULTS

Figure 2 Outgoing tide approximately 1 hour after high The Sea-RAI was successful in meeting its primary
tide mission objectives. The robot was deployed three times, two
on Dec. 18 and once on Dec. 19 from a sand spit about 1,000
Figure 3 is a screenshot of the Sea-RAI interface showing two ft from the bridge. It used waypoint navigation to travel to the
camera views of the bridge from the USV overlaid on a bridge and then was manually controlled near the bridge.
Google Earth imagery of the bridge from before Hurricane Ike. Missions lasted approximately 2 hours, with about 1 hour of
Rollover Pass illustrates the complexity and the amount of active investigation of the bridge and debris field in between
structures that may be present in littoral regions versus more changes in tides. The VideoRay and Ecomapper were also
common UMV operations in bays or the open ocean. deployed. The USV found no sign of scour or washout of the
bridge pilings, as can be seen in figure 5. The image on the
V. UNMANNED MARINE VEHICLES bottom shows the healthy pilings at the Rollover Pass Bridge.
They give off a bright, sharp line where the foot meets the
The three vehicles used at Rollover Pass are shown in Figure 4. ground. This is contrast to dark holes in the image on the top
The Sea-RAI unmanned surface vehicle is a custom platform showing the scour at pilings for a dock at Marco Island
based on two 6ft catamaran hulls, similar to Charlie13 but inspected after hurricane Wilma. The Sea-RAI also did not
more stable than the SCOUT.14 It is capable of autonomous find debris that would present a hazard to the bridge.
waypoint navigation and supports teleoperation. It is an
 During the Marco Island deployment the Sea-RAI The role of humans were piloting, serving as payload
predecessor was able to locate a collapsed section of the specialists (operating the sensors), serving as subject matter
bridge, Figure 6, and miscellaneous debris in the harborage experts (interpreting the data), and safety oversight. A
itself. minimum of four people 6 were involved. These roles relied
on a close interaction between all three groups. As noted
VII. GENERAL FINDINGS above, the robot was teleoperated for a significant portion of
the time and the robot was in line of sight for the entire
The evaluation and CWA produced findings in three mission.
areas: control challenges for UMVs, human robot interaction, A surprising human-robot interaction was the need
and uncertain data. The control and human-robot interaction for two-way audio, as there was always a person involved in
analysis confirmed earlier findings from Wilma and other launching or recovering the robot. At Rollover Pass, the base
CRASAR deployments plus observations from other marine station was 1000 ft from the launching area. Two way audio
vehicle research. The challenges in handling and fusing the would have permitted the operator to coordinate with the
sensor data are new. UMV control challenges. The fieldwork handler in the field rather than be reduced to gesturing at a
at Rollover Pass identified three major control challenges for
UMVs. The first is navigation in swift currents. The swift
currents limited the times and duration the Sea-RAI,
Ecomapper, and VideoRay could be used. The Sea-RAI was
actually put on a safety line during the first run to ensure it
could be recovered. The problems with currents and station
keeping is not unknown, see15 for another example. Second is
GPS loss or errors. As noted in,2 operations near bridges
interfere with GPS signals impacting the USV and the surface
operations of the UUV. The Sea-RAI had to be teleoperated
for the actual inspection task. GPS alone will not have

Figure 4 Sea-RAI (left), VideoRay ROV (Middle),

Ecomapper AUV (Right)

sufficient accuracy for close unaided inspection since GPS

errors in cluttered littoral environments can lead to collisions,
as seen when the Ecomapper while using GPS for a surface-
based scan of the channel bumped into a barricade. The third Figure 5 Top – DIDSON Imagery – Top, sign
challenge is obstacle avoidance for underwater vehicles. The of scour at Marco Island docks after Wilma.
Ecomapper could not safely operate submerged through the Bottom, no sign of scour at Rollover Pass after
channel until the Sea-RAI mapped the debris; unlike bays or Ike.
open water, littoral regions after a disaster may be cluttered camera and having the operator signal through panning or
with unmodeled obstacles. Also, the tethered VideoRay ROV tilting and having a spare team member run back and forth.
became tangled around a pipeline. This highlights the need for The need for multiple displays to accommodate multiple
tether management and awareness of where the tether is observers was noted in16 and the fieldwork reinforced this.
related relative to obstacles and the robot. Recent commercial Additional displays are needed for the specialists and subject
developments have created a “smart” tether which keeps up matter experts and the display should be customizable. For
with its own position, which might solve the problem in the example a civil engineer may just want to see the DIDSON
future. output while the operator sees the complete interface which
The fieldwork confirmed the role of humans in includes vehicle health, path, etc. The deployment also
shared autonomy, reinforced the need for multiple displays, showed problems with resilience and how hard it is for
and illustrated the lack of resilience in design and displays.
humans to understand what is going on with vehicles. The VIII CONCLUSION
DIDSON was knocked out of alignment by the force of the
water, causing the operators to get confused about which way The experience at the Rollover Pass Bridge showed
it was pointing relative to the vehicle. This led to coordination that Unmanned Marine Vehicles have sufficient utility for
challenges as the specialist had to give counter intuitive immediate use in littoral inspection. While tethered ROVs
commands to the pilot to maintain views of pilings. The have begun to be explored by transportation departments,
design of the payloads should prevent “normal” slippage and unmanned surface vehicles appear the most promising because
the display should provide diagnostics for confirmation of of navigability, ability to carry payloads such as acoustic
settings and positions. This has been addressed in the newest cameras which can penetrate turbidity, and real-time
version of the DIDSON visualization software, by providing transmission of data. Regardless of surface or underwater
an indication of the sonar orientation relative to the vessel it is deployments, UMVs pose many open research questions in
deployed on. control, human-robot interaction, cooperation between surface
and underwater vehicles, and handling of large data sets of
The UMV control challenges, particularly the uncertain sensor readings.
problems with station keeping and obstacle avoidance,
suggested that USV and UUV work cooperatively. The USV Acknowledgment
could map the region sufficiently for UUV navigation. As
noted in,2 the USV could also serve as a mother for a tethered This work was supported in part by a grant from the
ROV, seeing where the USV could not, reducing the amount Texas Transportation Institute. The authors thank Dr. Dennis
of tether and risk of tangling of a ROV, and help keep watch Christiansen (TTI) for his help, TEEX for the loan of response
on the tether. equipment, Daniel May for his expert help and suggestions,
The purpose of the UMVs is inspection, to observe and Kimberly Mallett for logistics and support.
and record; this creates challenges, particularly in handling
large datasets and managing uncertainty. The Sea-RAI References
collected a continuous stream of imagery from above and
below the waterline. The imagery was color video and sonar, [1] J. Brown, “Robot Inspects Damaged Bridge,” Civil Engineering,
leading to large datasets of information. Furthermore, the data Vol. 78, No. 3, 2009, p. 40.
had uncertainties in localization and in content due to shadows [2] R. Murphy, E. Steimle, E. Griffin, C. Cullins, M. Hall, and K.
and differing viewpoints from the vehicle and DIDSON angles, Pratt, “Cooperative Use of Unmanned Sea Surface and Micro Aerial
Vehicle at Hurricane Wilma,” Journal of Field Robotics, Vol. 25, No.
presenting challenges for accurate 3D reconstruction and in 3, 2008, pp. 164–180.
understanding. [3] R. Murphy, S. Tadokoro, D. Nardi, A. Jacoff, P. Fiorini, and A.
Erkmen, “Rescue Robotics,” Handbook of Robotics (B. Sciliano and
O. Khatib, eds.), pp. 1151–1174, Springer-Verlag, 2008.
[4] S.Williams and I. Mahon, “Simultaneous localisation and
mapping on the Great Barrier Reef,” Robotics and Automation, 2004.
Proceedings. ICRA ’04. 2004 IEEE International Conference on, Vol.
2, 2004, pp. 1771–1776 Vol.2.
[5] C. v. Alt, B. Allen, T. Austin, N. Forrester, R. Goldsborough, M.
Purcell, and R. Stokey, “Hunting for mines with REMUS: a high
performance, affordable, free swimming underwater robot,”
OCEANS, 2001. MTS/IEEE Conference and Exhibition, Vol. 1, 2001,
pp. 117–122 vol.1.
[6] G. J. S. Rae and S. E. Dunn, “On-line damage detection for
autonomous underwater vehicles,” Autonomous Underwater Vehicle
Technology, 1994. AUV ’94., Proceedings of the 1994 Symposium on,
1994, pp. 383–392.
[7] M. Caccia, R. Bono, G. Bruzzone, E. Spirandelli, G. Veruggio, A.
M. Stortini, and G. Capodaglio, “Sampling sea surfaces with
SESAMO: an autonomous craft for the study of sea-air interactions,”
Robotics Automation Magazine, IEEE, Vol. 12, No. 3, 2005, pp. 95–
105. 1070-9932.
[8] R. Schiavon, R. Finotello, and T. Terribile, “The supervisory
control of ARAMIS, a system for robotic inspection of sediments,”
OCEANS 2000 MTS/IEEE Conference and Exhibition, Vol. 2, 2000,
pp. 823– 828 vol.2.
[9] D. Ribas, P. Ridao, J. D. Tardos, and J. Neira, “Underwater
SLAM in a marina environment,” Intelligent Robots and Systems,
2007. IROS 2007. IEEE/RSJ International Conference on, 2007, pp.
Figure 6 Peir Section and Railing from Marco 1455–1460.
Island Bridge
[10] M. Lindemuth, R. Murphy, E. Steimle, W. Armitage, K. Dreger, [14] J. Curcio, J. Leonard, and A. Patrikalakis, “SCOUT - a low cost
T. Elliot, M. Hall, D. Kalyadin, J. Kramer, M. Palankar, K. Pratt, and autonomous surface platform for research in cooperative autonomy,”
C. Griffin, “Sea-RAI: A Marsupial USV-UAV Team for Littoral IEEE OCEANS, 2005, pp. 725 – 729.
Inspection,” IEEE Robotics and Automation Magazine, Vol. to [15] A. Pereira, J. Das, and G. S. Sukhatme, “An experimental study
appear, 2009. of station keeping on an underactuated ASV,” Intelligent Robots and
[11] M. D. Ageev, L. V. Kiselev, and A. P. Shcherbatyuk, “Tasks for Systems, 2008. IROS 2008. IEEE/RSJ International Conference on,
autonomous underwater robot,” Advanced Robotics, 1991. ’Robots in 2008, pp. 3164–3171.
Unstructured Environments’, 91 ICAR., Fifth International [16] R. Murphy and S. Stover, “Gaps Analysis for Rescue Robots,”
Conference on, 1991, pp. 1360–1364 vol.2. ANS 2006: Sharing Solutions for Emergencies and Hazardous
[12] R. Murphy, K. Pratt, and J. Burke, “Crew Roles and Operational Environments,2006.
Protocols for Rotary-Wing Micro- UAVs in Close Urban [17] M/ Hall and E. Steimle, “Unmanned surface vehicles as
Environments,” ACM SIGCHI/SIGART Human-Robot Interaction, Environmental monitoring and assessment tools”, Oceans 2006.
2008, pp. 73–80. [18] R. Murphy, “Trial by Fire,” IEEE Robotics and Automation
[13] M. Bibuli, G. Bruzzone, M. Caccia, G. Indiveri, and A. A. Magazine 2004. 11(3), 50-61.
Zizzari, “Line following guidance control: Application to the Charlie [19] R. Murphy, C. Griffin, S. Stover and K. Pratt, “Use of micro air
unmanned surface vehicle,” Intelligent Robots and Systems, 2008. vehicles at Hurricane Katrina” IEEE workshop on safety, security
IROS 2008. IEEE/RSJ International Conference on, 2008, pp. 3641– rescue robots, Gaithersburg. 2006.
3646.

View publication stats