January 25, 2015

Overview

On my first deployment as a United States Marine Corps Aviator with the 26th Expeditionary Unit, I was immediately inculcated with respect for the necessity of risk management while operating at sea. Within the first month of deployment during a long range airlift operation from ship to shore, I lost an engine on my dual engine CH-46E helicopter. The engine had failed after takeoff. The first thing that I thought of as looked down at the expanse of ocean between me and salvation back on the boat was, "thank God I can swim." But in hind sight I realize how lucky we were to have been able to limp back the USS Iwo Jima with one engine not only failed but smoldering in fire. We had been barely able to land within inches of the edge of the flight deck only by severely overtorquing the transmission of our over encumbered aircraft.
Later in the deployment my appreciation for my good fortune increased when three sailors fell into the Gulf of Aden while trying to deploy a raft from the deck of the USS San Antonio. They fell a distance of about 15 feet. Two of the sailors were immediately rescued and unfortunately the third was never recovered despite the entire episode being watched by several other sailors. After that I had a great respect for the expansiveness of the seas.

Search and Rescue (SAR) in the maritime environment involves considerable risk to SAR teams where survival times are short and minimum response time is crucial for recovery and survival of victims, see table (1). Unmanned systems are being developed to augment SAR. Unmanned systems can operate in adverse environments where low visibility and high sea states can place human rescuers at high risk unnecessarily. This article analyzes the operation of the ROAZ II that is being integrated into a search and rescue system developed by ICARUS and the European Union.
Table (1)

Water Temperature	Exhaustion or Unconsciousness in	Expected Survival Time
70–80° F (21–27° C)	3–12 hours	3 hours – indefinitely
60–70° F (16–21° C)	2–7 hours	2–40 hours
50–60° F (10–16° C)	1–2 hours	1–6 hours
40–50° F (4–10° C)	30–60 minutes	1–3 hours
32.5–40° F (0–4° C)	15–30 minutes	30–90 minutes
<32° F (<0° C)	Under 15 minutes	Under 15–45 minutes

(Cold Water Survival)

ROAZ II

As part of the ICARUS project, particularly in the maritime scenario, multiple heterogeneous unmanned platforms (by air or surface) will cooperate, in order to detect and assist castaways (Matos et al, 2013). One of those systems is the ROAZ II. ROAZ is a twin hull autonomous surface vehicle developed for oceanic robotics research and monitoring. It is based on a high density polyethylene 4.2m long catamaran capable of operating in ocean environment equipped with electric propulsion (two 2kW thrusters). With an autonomy of operation (depending on the battery configuration and speed) 8h to 10hr it can reach 10knots of maximum velocity (Matos et al, 2013). The ROAZ II was used as the delivery platform for an Unmanned Capsule which is capable of delivering and deploying a life raft with survival items to sea stranded victims.

Internal Systems

The internal systems are monitored by the supervising module [which] is responsible [for] [monitoring] a set of internal values, such as temperature, power consumption, or available energy, and send alert messages or trigger energy [sp] behaviors whenever faults are detected. It also monitors the communication link (Matos et al, 2013). This systems is comprised of a proprioceptive sensors that are all networks to the supervising module.

Navigation and Rescue

With the exception of the GPS receiver antennae autonomous navigation is primarily accomplished through a the use of proprioceptive sensors like a …GPS unit for absolute positioning (Novatel SmartAntenna, superstar II) and an IMU sensor coupled with magnetometer providing orientation, attitude velocities and accelerations. The IMU used is a Microstrain 3DM-GX1 module combining three angular rate gyros with three orthogonal accelerometers and three orthogonal magnetometers outputs orientation, angular rate and acceleration at a rate of more than 50Hz (Martins et al).
The ROAZ II also has the capability of using exteroceptive sensors for vision based target acquisition and tracking. ROAZ II [is] equipped with conventional cameras (both for on board image processing and for video transmission). ROAZ II system is equipped with a thermographic [sp] infrared camera (Fig. 5) capable of resolution up to 0.1ºC of temperature difference. The vision system processes [images] in real time, with edge detection and object identification, extracting target image characteristics (position, orientation) (Martins et al). Target position can be calculated into a 3D and 2D coordinate plane. A four state kalman filter is used to estimate target position and velocities from the raw vision data (Martins et al). From the target calculation the ROAZ II is able to maintain a fixed distance from the target.
Two other exteroceptive sensors, a side tracking radar provides bottom imaging capabilities and an external IEEE 802.11 a b/g Ethernet modem with external antenna which allows the ROAZ II to be operated remotely (Martins et al).
Once the ROAZ is within a nominal distance from its target under the ICARUS system it is then able to deploy a smaller Unmanned Capsule (UCAP) which is responsible for deploying a life raft and survival items within minimal distance to the stranded victim. The ROAZ is capable of holding multiple UCAPs and aid several victims.

Unmanned Aeronautical System (UAS) Integration

The ICARUS project is an emerging endeavor. As further tests are conducted other systems like UAS are to be integrated into the unmanned SAR system. UAS can travel faster and cover ground more efficiently than surface and underwater systems. For example, a UAS that operates below 10,000 feet typically can operate between 45kts and 200kts depending on aircraft design compared to a maritime surface system that typically operate between 15kts and 35kts also depending on the systems design. Integrating a UAS into the ICARUS system with ROAZ and UCAP will add an additional dimension and improve launch to the rescue times which are critical for survivability.

Improvements

Because the ROAZ is a twin hull design, in the event that the ROAZ is capsized it would be rendered ineffective. A counter balance system comprised of weights and self-inflating bags integrated into platform designed to pivot into an upright position in the event that it is capsized would allow the ROAZ to continue rescue missions despite being capsized.
An additional improvement would be the development and integration of a modular coordination system. This system could be designed so that multiple unmanned systems could take advantage of the full spectrum of space (underwater, surface, and air). Incorporation of multiple systems would further improve efficiency and response times as long as those systems are integrated in complementary organization.

Conclusion

Both manned and unmanned systems can be equipped with similar proprioceptive and exteroceptive sensors. However, the real advantages of using unmanned systems to augment manned SAR operations is the reduction of operating cost in both people and materials. The use of unmanned systems in adverse environments reduces the overall risk to the human rescuer as well as decreasing time to rescue for the victims. In short unmanned systems increase survivability.

References
Aníbal Matos, Eduardo Silva, Nuno Cruz, José Carlos Alves, Duarte Almeida, Miguel Pinto, Alfredo Martins, José Almeida, Diogo Machado. Development of an Unmanned Capsule for LargeScale Maritime Search and Rescue. (2013). Retrieved on January 25, 2015 from http://oceansys.fe.up.pt/publications/2013_MatosSilvaCruzAlvesAlmeida.pdf

Cold Water Survival. (n.d.). Retrieved on January 25, 2015 from http://www.ussartf.org/cold_water_survival.htm

Alfredo Martins, Hugo Ferreira, Carlos Almeida, Hugo Silva, José Miguel Almeida, Eduardo Silva. ROAZ and ROAZ II Autonomous Surface Vehicle Design and Implementation. (n.d.) Retrieved on January 25, 2015 from www.researchgate.net/.../224982209_ROAZ_and_ROAZ_II_

Inertial Navigation Systems (INS) use the acceleration detected by sensors like lasers, accelerometers, gyroscopes, etc. and algorithmic equations to calculate their position relative to the frame of the Earth by measuring the centrifugal force from the rotation of the Earth. The underwater environment increases the complexity of calibrating an INS when compare to systems used in aviation since aviation INSs drift errors are typically updated by the onboard Global Positioning System (GPS). However, INS is not dependent on GPS in order to function proper. GPS merely helps eliminate INS drift error which is inherent to modern systems. In contrast, INSs that operate in underwater environments are only able to utilize GPS corrections when the vehicle is at the surface. The article Achieving High Navigation Accuracy Using Inertial Navigation Systems in Autonomous Underwater Vehicles by Robert Panish and Mikell Taylor for the Bluefin Robotics Corporations demonstrates the calibration methods of two different INS systems, the T-24 Ring Laser Gyro (RGL) and PHINS III Fiber Optic Gyro (FOG), which are both used on their BlueFin Autonomous Underwater Vehicle. The operational advantages of each system is minimal in comparison through the evaluation of each INS calibration method presented in this article.

The T-24 uses RLGs, which means the beam path is created by a set of mirrors redirecting the laser into a loop (Panish and Taylor), while the PHINS III, the laser beam travels through a long optical fiber to create the beam path (Parnish). Both systems measure the change of the laser’s frequency as a result of the laser being bent by acceleration forces outside of the system. The measured change in frequency allows the INS to calculate its linear acceleration. When this linear acceleration is referenced with the Earth’s rotation the INS is able to calculate the systems location by the corresponding it to a unique acceleration vector on each point of the Earth. Each of these systems uses of a Dopper Velocity Log (DVL), depth sensor and sound speed log (Panish) with the INS in order to calculate its position over time and eliminate drift error. And both systems use GPS corrections during calibration in order to minimize the initial drift error. However, the actual methods of each is calibration is unique.

During the calibration of the PHINS III INS, inertial sensors and GPS are used to determine its motion. The DVL velocities are recorded. The difference in known motion from the INS and GPS in comparison to the DVL motion determines its calibration parameters. This is done by sending the vehicle on a 5km track line with continuous GPS contact and monitoring the convergence of roll, pitch, and heading misalignment angles (Panish). Dissimilarly, during the calibration procedure of the T-24 INS the vehicle is submerged and follows a box shaped pattern with surfacing at each of the corners. Each side takes no less than fifteen minutes. Using the GPS fixes at each corner it determines its internal biases, scale factors, and misalignment angles (Panish). Roll and pitch are determined from a simple measurement of the direction of the gravity vector from the accelerometers measurements. While heading is determined by using the time derivative of the gravity vector, easterly, and the Gravity vector in order to calculate north. While the alignment of the PHINS III took less time to calibrate than the T-24, the PHINS III was more susceptible to sea state and required more monitoring because of the possibility of collision with other surface ships during calibration.

Although both of these methods measure accelerations differently, both of these methods calibrated the two INS systems within a drift error less 0.1% of distance traveled. The drift error of these two systems far exceeds the design specifications for the system. The minimization of drift error during calibration is important when evaluating these systems since it will not be able to utilize the GPS to correct for drift while being submerged for long durations and distances. Since the most notable difference in an operational point of view is the calibration method, each of these systems provides exceptional navigation accuracy that can be used to collect high quality oceanographic data (Panish).

Robert Panish and Mikell Taylor. Achieving High Navigation Accuracy Using Inertial Navigation Systems in Autonomous Underwater Vehicles. (2011) Retrieved January 18, 2015, from http://www.bluefinrobotics.com/news-and-downloads/papers-and-articles/

Sunday, January 25, 2015

Unmanned Maritime Search and Rescue, ROAZ II