Development and Operational Experiences of UAVs for Scientific Research in Antarctica

Abstract

Several models of UAVs have been developed for scientific missions such as aerial geomagnetic survey, meteorological observation, and ground observation instead of manned airplanes especially in Antarctica. The UAV fleet is divided into three groups, i.e., a small UAV for low altitude, relatively large UAVs for mid-altitude, and a small glider UAV for high altitude. Several results, operational experiences of the UAVs, lessons learned including the operations in Antarctica, and several problems to be solved are reported in this chapter.

1 Introduction

The authors have been developing several unmanned aerial vehicles (UAVs) and trying to use them as tools for scientific researches especially in Antarctica. Small manned airplanes had been operated for various scientific missions at Syowa Station in Antarctica for Japanese Antarctic Research Expedition (JARE) team until 2005, but the operation ended after 2005 mainly due to its high operational cost and several difficulties in severe environment at Antarctica. It triggered the development of our UAVs which can be used for scientific missions instead of manned airplanes, and some of the UAVs have begun to be used in Antarctica.

UAVs can be used for several scientific missions such as aerial geomagnetic survey, meteorological measurement, biological observation, and glaciological observation. Geological structure under the ground even covered by ice can be estimated by examining geomagnetic anomaly over wide area by flying comprehensively on a certain surface at some altitude. Precious meteorological data in 3D space can be obtained by climbing up to necessary altitude and by flying over necessary area. Biological and glaciological information can be acquired by taking pictures and movies from the UAV. Such image information is also quite useful for the ground expedition team in planning for expeditions. The most advantageous points using UAVs compared with using manned airplanes for such missions are low cost and mitigation of pilot’s safety issues. Mitigation of pilot’s safety issues is especially important in severer environment like in Antarctica.

On the other hand, there are some different problems to be solved not only in technical area but also in operational aspect for using UAVs in such non-normal environment. Low temperature environment causes several problems such as deterioration of battery performance and structural fragileness of some materials used in UAVs. Wind speed in Antarctica changes depending on the area and season, but strong wind blows most of the time in some area, and it makes difficult to operate UAVs especially in takeoff and landing. Logistics is another issue which cannot be bypassed because there is always limit for the amount of goods which can be carried for the operation and it is almost impossible to supply necessary parts or goods immediately as required especially in Antarctica.

In this chapter, details of the UAVs developed and operational results and experiences and lessons learned from the experiences are described.

2 Development of the UAVs

Major missions required for the UAV are aerial geomagnetic survey, meteorological survey, and ground observation for various purposes. These missions require different capabilities of the UAV. In order to perform aerial geomagnetic survey, it is required to fly at relatively low speed and relatively low constant altitude over wide area while the weight of the measurement apparatus is not so heavy (<1 kg). For meteorological survey, ample excess power is required in order to climb up to high altitude of interest. It could be also required to fly over wide area depending on the scale of the meteorological phenomena, and some of the measurement apparatuses such as an optical particle counter and an aerosol sampler tend to be relatively heavy (2–3 kg in total). For ground observation, major payload is a digital camera or movie camera, and its weight is <1 kg, but some anti-vibration equipment is necessary for taking clear images or movies. From these considerations, it is thought to be difficult to meet all these requirements at once by one vehicle, and two types (three models) of UAVs have been developed or under development depending on the mission.

One type is an engine-driven UAV [1], and there are two models named as AntPlane-3 and AntPlane-6. The word “ant” comes from “Antarctica” and an “ant” as an insect because the UAV is very small compared with a manned airplanes. AntPlane-3 is a relatively small UAV, with a 20-cm³ four-cycle gasoline engine, and suited for aerial geomagnetic survey and ground observation at low altitude. AntPlane-6 is larger than AntPlane-3 equipped with an 86-cm³ two-cycle gasoline engine, and it can be used even for meteorological missions up to middle altitude.

Another type is a motor glider with an electric motor combined with a weather balloon. It is currently under development and flight testing. We call it balloon-assisted UAV. It is planned that the balloon-assisted UAV climbs up to the altitude around tropopause using buoyancy of the balloon. Sampling and density measurement of so-called aerosol will be performed during its ascent, and the function of the UAV is recovery of the measurement apparatuses onboard. “Aerosol” is quite small particles drifting in the air. It becomes nuclei of rain drops, and it is said that aerosol is closely related to various climatic phenomena and plays important roles. After reaching some planned altitude, the UAV will be separated from the balloon and fly back to the released point on the ground mainly by gliding. It could use thrust provided by an electric motor in the case that the UAV reaches close to the ground before returning to the released point.

2.1 AntPlane-3

A picture of AntPlane-3 is shown in Fig. 11.1, and its specifications are shown in Table 11.1. Airframe of the AntPlane-3 is modified from off-the-shelf RC model airplane of the conventional form, and the development cost becomes very low. Its fuselage is made of GFRP, and the wings are made of styrene foam core covered with thin wood plate and film. A four-cycle 20-cm³ gasoline engine is also an off-the-shelf product. Major limiting factor of range is of course the quantity of fuel, and the next is the capacity of battery for all electric systems onboard. As the capacity of a battery increases, unnecessary weight increases and increasing time required for charging is also wasteful. An engine-driven generator has been developed using a brushless electric motor for RC airplanes, and it provides all electric power required for the whole system including an automatic flight control module while charging a five-cell Ni–MH battery pack. Details of the automatic flight control module will be described later in this section. This system can provide electric power for a while even after fuel becomes empty until the battery becomes empty. Takeoff and landing are done manually by a human pilot, and the flight mode is switched to automatic flight mode after takeoff. It is equipped with emergency parachute recovery system in the canopy. The parachute can be deployed manually, automatically by the automatic flight controller specifying a certain condition, and also automatically by mechanical system even all electric power is out. The system has been flight tested in Japan and confirmed its proper function.

Fig. 11.1

AntPlane-3 under preflight check at Marsh Airfield, King George Island, Antarctica

Table 11.1 Specifications of AntPlane-3

A magnetic sensor for magnetic field survey is mounted in a pod at a wing tip, and the closest servo motor to the magnetometer is wrapped by a piece of magnetic shield film in order to avoid the unnecessary change of magnetic field due to servo motor movement. A digital camera is also mounted in a pod at another wing tip.

2.2 AntPlane-6

A picture of AntPlane-6 and its specifications are shown in Fig. 11.2 and Table 11.2, respectively.

Fig. 11.2

AntPlane-6 under preflight check at Marsh Airfield, King George Island, Antarctica

Table 11.2 Specifications of AntPlane-6

Different from AntPlane-3, its airframe is made by an RC modeler based on our requirements of the specifications. All parts including fuselage and wings are made by sandwich structure composed of a thin styrene foam plate and GFRP layer using molds. A two-cycle 86-cm³ gasoline engine is mounted at the short rear end of the fuselage, and the horizontal tail is supported by twin boom as shown in Fig. 11.2. Its fuselage is composed of front payload module and aft module with all parts and systems necessary for flight. A user can replace front payload module depending on their desirable mission. Figure 11.2 shows that a magnetic measurement module is used as the front payload module. A magnetic sensor is mounted at the tip of nose boom in order to avoid magnetic effects of the onboard system, and a data logger is mounted in the fuselage part of the front payload module. If a camera module is connected, not only a high-resolution digital camera but also a Hi-Vision digital movie camera can be used. AntPlane-6 is similarly equipped with the same automatic flight control module as in AntPlane-3, emergency parachute recovery system in its midwing section, and a generator. The whole system has been also tested in Japan and confirmed its proper function.

2.3 Balloon-Assisted UAV

A picture of balloon-assisted UAV (prototype[2–5]) and its specifications are shown in Fig. 11.3 and Table 11.3, respectively. Basic airframe of the balloon-assisted UAV is the same as AntPlane-3, but the powerplant is replaced by an electric motor with folding propeller. The propeller is folded down along the surface of the nose for drag reduction during its glide.

Fig. 11.3

Ascending balloon-assisted UAV (prototype test flight)

Table 11.3 Specifications of balloon-separation UAV

The UAV is connected to a helium-filled weather balloon (diameter at sea level: 3m) with a piece of very thin synthetic fiber string called Dyneema. The string is cut off at planned altitude by small heaters (we call it “separator”) mounted in aft fuselage and actuated by the automatic flight controller at two points of the string for redundancy. The balloon is also equipped with an additional off-the-shelf “separator” for weather balloons in case of the malfunction of the two separators of the UAV.

For direct observation of aerosol, an optical particle counter (OPC, Fig. 11.4) and aerosol sampler (Fig. 11.5) are carried in the fuselage of the UAV. Downsizing of these measurement apparatus is current issue for the realization. An apparatus called a GPS sonde and a data transmitter (Fig. 11.6) is also mounted on the UAV. The GPS sonde measures basic climatic data such as air temperature and humidity with 3D position information acquired by a GPS receiver. The 400-MHz transmitter transmits these data to the ground station. The data obtained by the optical particle counter is also transmitted via the same downlink system, while aerosol samples have to be carried back to the ground by the UAV.

Fig. 11.4

An optical particle counter (OPC) for aerosol density measurement

Fig. 11.5

An aerosol sampler

Fig. 11.6

GPS sonde sensor and transmitter

2.4 Automatic Flight Control Module

An automatic flight control module [6] has been developed in-house combining two microcontroller boards, a sensor board, and a GPS receiver as shown in Fig. 11.7.

Fig. 11.7

(a) A microcontroller board and a sensor board, (b) an automatic flight control module

An H8-2638(20 MHz) microcontroller of Renesas Electronics is mounted on a microcontroller board (Fig. 11.7a, left). The board has a form factor of 3 cm × 9 cm. A sensor board (Fig. 11.7a, right) has the same form factor, and MEMS sensors such as differential pressure sensor, an absolute pressure sensor, 3-axis angular rate sensors, and 3-axis accelerometers are mounted with anti-aliasing filters.

All the parts composing the module are chosen in order to meet the supposed temperature range requirement from −25 °C to 80 °C, and the module function is tested in several environments at different temperatures. Previously described all three UAVs are equipped with this control module.

It is possible to switch servomotor driving signals between an RC receiver (manual flight mode) and the output of the controller (automatic flight mode) by a human pilot on the ground using an RC transmitter. Usually, the UAV takes off by manual flight mode, and then it is switched to automatic flight mode by the pilot. It is switched back to manual flight mode before landing, and then the pilot makes the UAV land manually.

In automatic flight mode, airspeed and altitude are controlled by PID control law in longitudinal motion. Course over ground control and side-slip suppression is performed also by PID control law in lateral motion. Airspeed command, altitude command, and waypoint positions are specified by the ground station software, and the course command is generated according to the relationship between current position and next waypoint.

In the onboard software for the balloon-assisted UAV, launch mode, pull-up mode, and automatic glide mode are added. In launch mode, all control surfaces are kept neutral and the electric motor is kept stopped. Altitude is always watched for separation. Once the UAV reached the specified altitude, the separator activates and cuts the string for connection between the UAV and the weather balloon. Pull-up mode pulls up the nose of the UAV and is followed by glider mode in which only airspeed is controlled in longitudinal motion. The airspeed command which gives best glide ratio is determined by estimated wind speed and the polar curve of the UAV. Lateral control exactly the same way as in the automatic flight mode is performed in the glider mode.

2.5 Ground Station Software

Ground station is composed of a laptop computer and a wireless modem. Waypoints information with airspeed and altitude commands in between the waypoints have to be specified by the ground station software before takeoff. During flight, current position of the UAV is displayed on a map as shown in Fig. 11.8.

Fig. 11.8

Current position of the UAV displayed in the ground station software

All sensor data can be monitored, and all control parameters can be adjusted also by the ground station software as shown in Fig. 11.9.

Fig. 11.9

Data monitor window and control parameter input window

3 Operational Experiences and Lessons Learned

We used to own UAVs named AntPlane-1 and AntPlane-2 [7] developed by the cooperation of Nippi Corporation, and AntPlane-4 developed by the cooperation of Fuji Imvac Inc. other than previously described UAVs such as AntPlane-3, 6, and the balloon-assisted UAV. AntPlane-1, 2, and 4 were almost in similar size as AntPlane-6 and equipped with also similar size engine as AntPlane-6. AntPlane-1 and 2 are tractor-type UAVs, and AntPlane-4 is a pusher-type UAV.

Several attempts to fly the UAVs and to acquire geomagnetic and meteorological data not only in Japan but also in foreign countries and Antarctica have been made [8]. At Syowa Station in Antarctica, several flights have been attempted by AntPlane-1 in 2006 as one of the activities of JARE 46th (JARE46), by AntPlane-4 in 2007 (JARE47) and 2008 (JARE48), but every attempts were not successful unfortunately. Major causes are human error due to insufficient operational training and engine failure due to environmental difference. In every attempt, one of the scientists in expedition team has received training in short time before leaving Japan in summer, but the training was not sufficient. Flight environment, especially air temperature and humidity, is quite different during training in summer in Japan and even summer in Antarctica. Sufficient training and sufficient flight tests in Japan under similar conditions as in Antarctica are strongly required.

One example which has shown the significance of UAVs is the geomagnetic survey flight performed in Kalgoorlie, Australia, in 2006. An AntPlane-4 was used, and the flight was done by the support of Fuji Imvac Inc. The UAV flew over the area of 10 km × 10 km, along a straight line spaced every 250 m at an altitude of 500 m above ground level as shown in Fig. 11.10. Total flight distance was 505.7 km, and the flight took 3 h and 36 min.

Fig. 11.10

Flight track of the AntPlane-4 in geomagnetic survey flight at Kalgoorlie, Australia, in 2006

Measured geomagnetic anomaly by the UAV is shown in Fig. 11.11. Geomagnetic anomaly corresponds to the same area measured by a manned airplane and publish by Australia Geoscience is shown in Fig. 11.12. Comparing these two figures, geomagnetic anomaly pattern of the corresponding area is quite similar, and this means the measurement by the UAV is useful. It is also significant from the viewpoint of cost. The measurement by the UAV is much cheaper than using a manned airplane.

Fig. 11.11

Magnetic anomaly measured by AntPlane-4 at Kalgoorlie, Australia, in 2006

Fig. 11.12

Magnetic anomaly measured by a manned airplane and published by Australia Geoscience

Another significant example of UAV flight has been done at King George Island, Antarctica, in 2011. Geomagnetic survey flights are planned around King George Island in Antarctica and Bransfield Strait using AntPlane-3 and -6. Planned survey area around King George Island is shown in Fig. 11.13. Forming mechanism of Bransfield basin is expected to be investigated by geomagnetic anomaly survey. One of the climatic features around there is ever-blowing strong wind (average 10 m/s). Our team consists of two scientists, one engineering researcher, and two cooperators who have stayed at Chilean Escudero Station for about 1 month and waited for calm weather assuming to use the runway of Chilean Marsh Airfield. Some of the difficulties in addition to the strong wind are fog on calm weather conditions, limitation of runway use due to manned airplane operation on fine days, limitation due to the operating hours of Marsh tower, and ATC (air traffic control) communications required by the Marsh tower even no manned airplane is in operation.

Fig. 11.13

Planned flight area around King George Island in Antarctica

Due to these limitations and adverse weather conditions, we had very limited chance of UAV flights. Figure 11.14 shows one of the flights of AntPlane-6 for UAV system test and the magnetic measurement system test on a miraculous fine day (wind speed on the ground: 3 m/s, air temperature: 5°C) over Marsh Airfield flying along a square track.

Fig. 11.14

Flight track over Marsh Airfield at King George Island in Antarctica

Flight track which deviates considerably from the lines connecting four waypoints is the track in manual flight mode, and the flight track in automatic flight mode shows good results. The measured geomagnetic change during this flight is shown in Fig. 11.15. Geographical feature around the area is homogeneous basalt, and the change in magnetic strength in Fig. 11.15 is considered to correspond to the geographical feature (mountainous terrain in east area, and valley and sea in west area). It shows that the aerial geomagnetic system is useful.

Fig. 11.15

Measured geomagnetic anomaly over the flight track at King George Island in Antarctica

Although we were not able to perform scientifically meaningful survey flights this time, the fact that the UAVs developed by ourselves could be operated by the team mainly composed of scientists and researchers is significant. The survey flights will be performed again from the end of November 2011 to mid-January 2012, and scientifically significant data are expected to be obtained by this operation.

4 Conclusion

Several models of UAVs that the authors have been developing are described. The UAVs are intended to be used for scientific missions such as aerial geomagnetic survey, meteorological observation, and ground observation instead of manned airplanes. The UAV fleet is divided into three groups, i.e., a small UAV for low altitude (AntPlane-3), relatively large UAVs for mid-altitude (AntPlane-1, -2, -4, and -6), and a small glider UAV for high altitude (balloon-assisted UAV).

Although scientifically significant data have not been obtained yet up to this moment, scientific survey flights using UAVs instead of manned airplane are certainly progressing its steps. On the other hand, it is true that there are still many problems to be solved for easy and reliable operation of UAVs such as system reliability, training and preparation issues, and weather issues.