You are on page 1of 4

061215-048

Tests of a light UAV for naval surveillance


A. M. Gonalves-Coelho, Lus C. Veloso, and Victor J. A. S. Lobo, Member, IEEE
such a wingspan, the UAVs maximum take-off weight should not exceed 6 kg. For a matter of high impact strength, as required for very hard and crash landings, the UAV structure should be made of expanded polypropylene (EPP). As a bonus, the EPP structure should provide improved buoyancy for the case of a crash landing at sea. For this small sized aircraft, the power plant should be a two-stroke, glow-plug internal combustion engine, because this type of engines delivers a substantial amount of energy at a relatively small payload. In fact, one of the final goals of the project is to achieve an operational flight range of 60 km (32 nmi), which was found to be suitable for a typical mission. Several papers have been written over the years about UAVs. However, not a single document was found about UAVs with the required characteristics. Therefore, some inspiration was drawn from the Raven and the Dragon Eye light UAVs [1]. Although they are powered by electric motors, the size and weight of both the Raven and the Dragon Eye is of the same order of magnitude. The take-off of our UAV could be carried out by hand launching at the forecastle. As for the landing, the sole eligible part of the ship is the stern section, a very small and irregular area with a size of around 5x6 m. This section is mostly taken by a bay for the operation of a rubber dinghy, as shown in Fig. 1.

Abstract The present paper describes the current state of the ongoing project of an unmanned aerial vehicle (UAV) intended for surveillance of territorial waters. The UAV will operate from the Portuguese Navy Argos and Centauro class fast patrol boats (FPB), which implies the capability of taking-off and landing in 27 m length, 5.9 m breadth vessels with a small and irregular landing zone at the boats stern, with an area around 5x6 m. In the final version, the UAV will have a 12 nm range, and the ability to perform GPS supported autonomous flight, automatic take-off and landing, as well as georeferenced imaging with realtime radio transmission to the FPB. Index Terms Unmanned aerial vehicle, naval surveillance.

I. INTRODUCTION HE Portuguese coastline including Madeira and Azores islands is 1860 km (1000 nmi) long. This makes it difficult to control the approximately 41335 km2 territorial waters, a duty in which the fast patrol boats (FPB) of the Portuguese Navy have a remarkable share. The efficiency of those FPBs could be improved if their regular operation could be supported by unmanned aerial vehicles (UAV), with basic electronic surveillance capability i.e., able to carry out georeferenced imaging with real-time radio transmission to the FPB. Due to the specific operating environment, the adopted UAV should be simple, reliable, small and strong. Moreover, it should be low-priced, due to the required number of units and to its intrinsically high operational risk. In fact, beside the ability for operating under severe weather conditions, the UAV should be able to be deployed and recovered onboard Argos Class FPBs, which are 27 m long, 5.9 m in breadth, draft 2,8 m, and displace 97 tonne. Centauro Class FPBs (28,4 m long, 5.95 m in breadth draft 2,8 m, and displace 98 tonne) were considered as well. Therefore, the project started with the search for the most appropriate solution for the above-mentioned requirements. The need for mechanical simplicity, reliability and cost effectiveness lead us to the selection of a fixed wing aircraft solution. However, the UAV wingspan should not exceed 1.8 m in order to allow its handling both inboard (for storage and maintenance) and at the take-off and landing zones. With
This work was supported by the Portuguese Naval Academy. A.M. Gonalves-Coelho (PhD, Eng.) is with the Faculty of Science and Technology, The New University of Lisbon, Portugal. Lus C. Veloso (BSc) is with the Portuguese Naval Academy, Alfeite, Portugal. Victor J.A.S. Lobo (PhD, MSc, Eng.) is with the Portuguese Naval Academy, Alfeite, Portugal.

Fig. 1. The stern section of a Centauro Class FPB with its bay dedicated to the operation of a rubber dinghy. The net that is located at the top of the photo was set up to provide a landing zone for the UAV.

For this reason, different concepts for the landing support system were evaluated, and it was found that the best system should be a nylon cord net. In fact, some preliminary tests that

061215-048 where carried out in terra firma with different types of stopping systems, such as horizontal and vertical stop cables, and different types of nets, have shown that a net made of nylon cord is able to sustain very hard crash landings without noticeable damage of the UAV. In addition, Fig.1 shows that a net fixed to the vessels superstructure does not disturb any essential function of the FPB. More important, the dinghy operation is not disrupted. The survival of the complete system under very hard operating conditions is a crucial point for the ongoing project, and the present paper is an attempt to show the appropriateness of the main design decisions that have been made on this particular subject.

2 The typical take-off weight for this configuration is approximately 3 kg, which is well below our 6 kg first guess.

II. THE COMPLETE EXPERIMENTAL SETUP A commercially available 1.72 m wingspan, radiocontrolled aircraft model with EPP structure has been selected as a means to proof the adopted concepts (Fig. 2).

Fig. 3. The UAV after a crash landing at the sea. Note the good buoyancy that characterizes the EPP structure.

Fig. 4. One of the video cameras that were used in some flight tests.

Fig. 2. The UAV prototype as used in the flight tests at the sea.

The aircraft was carefully prepared for possible crash landings at sea, with all the electronic equipments made watertight by means of rubber balloons, lubricant grease and glue. The prototype proved to be waterproof after being submerged in a water tank for a 24-hour period. Moreover, it has flown without any problem immediately after a crash landing at the sea (Fig. 3). Some active flight stabilization systems have been tested as well, such as an inertial platform [2] and a system that senses the difference in infrared signature between the earth and the carbon dioxide in the atmosphere [3]. The former system had produced the best results. The prototype was equipped with one or two video cameras for some of the flight tests. A radio link between the UAV and the vessel was set up to allow real-time video acquisition, and Fig.s 4, 5 show some details of the used equipment.

Fig. 5. The video acquisition system that was used onboard the FPB.

On the FPB side, hand launching from the forecastle deck was used, therefore no special provisions where required for the take-off. As for the landing, a football-goal type net made of nylon cord that was fixed to the FPBs superstructure to make available the UAVs landing zone (Fig. 6).

061215-048

Fig. 8. The UAV in the final approach for landing on NRP Centauro. Fig. 6. Installing the stopping net at the stern section of the vessel.

III. THE FLIGHT TESTS AND THE RESULTS Several test sorties have been done, the first four ashore, at different locations, followed by 5 sorties at sea, onboard NRP Centauro, NRP Cassiopeia, and NRP Hidra. The purpose of the shore tests was to understand the UAVs flight behavior and to test the eligible stopping systems in a much more favorable situation than the onboard conditions. A typical flight test on the sea begins with hand launching the UAV from the FPBs forecastle (Fig. 7) and ends with a very hard landing at the stopping net located at the stern of the ship (Fig. 8, 9).

As for the FPB, different cruising speeds were used, ranging from 0 to 47 km/h (0 to 25 knot), in increments of 5 knot. Fig. 9 illustrates a typical onboard landing.

Fig. 9. The UAV over the stopping net immediately after landing. Note that some fixing points of the net are made of elastic straps to improve elastic deformability.

The flight tests were performed along a 12-month period, with very different conditions, from sunny weather and almost still air to heavy raining with strong wind and 33 knot gusts.

Fig. 7. Hand launching the UAV at the forecastle of a FPB.

Typically, it follows a short flight around the FPB since the main objective was to evaluate the behavior of the complete experimental take-off and landing setup. Some longer flights of 10 to 15 minutes were made to test the active flight stabilization systems and the video cameras. Fig. 8 shows the UAV during a final approach. The speed of the UAV relative to the vessel was continuously monitored during the final approach by using a police type hand-held Doppler radar. The range of the landing speed has been 30 50 km/h (16 27 knot).

Fig. 9. NRP Hidra, as captured by one of the UAVs video cameras during a test flight over the Tagus estuary.

No attempts were made to check for the operational range of the UAV prototype since the main goal of the tests was to appraise the solutions that were found for taking-off and landing. Fig. 9 and 10 contain frames of two video streams that were collected during the flight tests.

061215-048

Fig. 10. An aerial view of the Portuguese Naval Academy taken by the UAV.

Over the course of our experiments, approximately 100 successful onboard landing were accomplished. Although in the first tests we did have some crash landings in the water, during the last sorties, a 100% success rate was attained, leading us to believe that the system is now robust and reliable. IV. CONCLUDING REMARKS The goal of this system is to design and build a UAV with a 32 nmi operational range, capable of taking-off and landing on Portuguese Navy Argos and Centauro class fast patrol boats or bigger vessels, that can perform autonomous flight with automatic take-off and landing. Autonomous flight will be achieved by using a GPS, an active flight stabilization system, pressure sensors and a stereo vision system. Tests so far have proved that this system is viable, and one of the most critical aspects (landing on such a small vessel) was proved possible.

ACKNOWLEDGMENT The authors would like to thank the cadets that helped in all the test program, as well as in many of the design and construction tasks. We would also like to acknowledge the support given by the Portuguese Naval Command, the Captains and crews of the vessels involved, and the Naval Base Police service. A special mention is due to our test pilots that always showed great enthusiasm and remarkable skills. REFERENCES
Goodman Jr., G.W., Congested airspace, C4ISR Journal, Vol. 5, No. 1, pp. 18-21, Jan/Feb 2006. [2] GY Series Gyroscopes. Available: http://www.futabarc.com/radioaccys/futm0807.html [3] Co-pilot stabilization system. Available: http://www.fmadirect.com/detail.htm?item=1489&section=20 [1]

You might also like