Unmanned aerial systems (UAS) are associated with a host of terms, reflecting the variety of existing configurations and possible fields of application. An unmanned aerial vehicle (UAV) flies either remotely and fully controlled from another place (e.g. ground, another aircraft) or programmed and fully autonomous (ICAO 2011). An UAV or UAS comprises the flying platform, an aircraft designed to operate without human pilot on board; the elements necessary to enable and control navigation, including taxiing, take-off and launch, flight and recovery/landing; and the elements to accomplish the mission objectives: sensors and equipment for data acquisition and transfer of data—including devices for precise location when necessary.
Aerial and remotely controlled systems for surveillance and acquisition of Earth surface data have a relatively long history, typically associated with military activities. Photogrammetry and remote sensing technologies identified the potential of UAV/UAS sourced imagery acquired at low altitudes with high spatial resolution, more than 30 years ago (Colomina and Molina 2014). However, civilian research on UAVs only began in the 1990s (Skrypietz 2012). Currently, a profuse emergence of UAV in civilian applications’ domains (e.g. agriculture, forestry, mining, marketing, patrolling, habitat, viticulture) has raised awareness of the potential of these aerial systems (a comprehensive review of environmental applications using UAVs can be found in Pajares 2015).
UAVs are classified under different schemes, using criteria such as flying height and range, size and weight (frequently referred to as maximum take-off-weight—MTOW). A strict categorization of UAVs is not however possible because certain characteristics in the various classes overlap (Skrypietz 2012).
The very small platforms, micro and mini aerial vehicles, can fly for less than 1 h at an altitude below 250 m. Micro platforms are considerably smaller than mini platforms (i.e. <5 versus 20–150 kg) but both have a similar flying range. Mini is the most abundant type of platform produced for civilian applications, doubling the number of micro and medium range UAV platforms (UVS 2014). An example of mini UAV is the Camcopter, with an MTOW of 68 kg and maximum payload capacity of 25 kg. On the other end of the scale, medium altitude long endurance (MALE) platforms (e.g. Talarion, Predator) and high altitude long endurance (HALE) platforms (e.g. Global Hawk, Euro Hawk) have a flying endurance of several days at an altitude up to 8000 and 20,000 m, respectively. The latter aerial platforms are comparable in size to manned aircraft. Developments of the technology are now providing nanodrones, miniature UAVs able to carry small still and video cameras. These UAVs can fly in all directions and perform manoeuvres and mid-air stunts. For example, the palm-size Micro Drone 2 weighs 0.034 kg and has a flying range of 120 m and endurance of 6–8 min. Other small drones are now flown as tethered aerial vehicles to circumvent the risks associated with flying. The Pocket Flyer by CyPhy Works is a 0.080-kg tethered platform that can fly continuously for 2 h or more, sending back high quality HD video the entire time. With improved tether technology, all data, control and endurance can be built into the tether, providing long endurance. Furthermore, ZANO operates on a virtual tether connected to a smart device, allowing simple gestures to control it. Groups of small UAVs deployed in formation with the same mission and intercommunicated form a swarm. Extensive literature about UAV configuration has lately emerged (e.g. Austin 2010; Valavanis and Vachtsevanos 2015). A brief and non-exhaustive description of the main elements (platforms, sensors and auxiliary equipment) now follows.
Platforms
Small UAV platforms are typically grouped into two main categories: rotary wing UAVs and fixed-wing UAVs. The capacity for vertical take-off and landing (VTOL) as opposed to horizontal take-off and landing (HTOL) was a valid criterion for categorizing UAVs, until some fixed-wing airframes also acquired VTOL capacity. Fixed-wing UAVs have a relatively simple structure making them stable platforms easy to control during autonomous flights. Efficient aerodynamics enables longer flight duration and higher speeds, which make them ideal for applications such as aerial survey requiring the capture of geo-referenced imagery over large areas. On the down side, fixed-wing UAVs need to fly forward continuously and need space to both turn and land. These platforms are also dependent on a launcher (human or mechanical) or a runway to facilitate take-off and landing, which can have implications on the type of payloads they carry. Typical lightweight fixed-wing UAVs currently in the commercial arena have a flying wing design with wings spanning between 0.8 and 1.2 m and a very small fin at both ends of the wing. In-house vehicles tend to have slightly longer wings to enable carrying the required heavier sensors (Petrie 2013). A second type of design is the conventional fuselage, with dimensions around 1.2 to 1.4 m length for the fuselage and 1.6 to 2.8 m wing length.
Rotary wing aircraft (multicopter or multirotor) have more complex mechanics, which translates into lower speeds and shorter flight ranges. Amongst their strengths, rotary wing UAVs can fly vertically, take-off and land in a very small space, and can hover over a fixed position and at a given height. This makes rotary wing UAVs well suited for applications that require manoeuvring in tight spaces and the ability to focus on a single target for extended periods (e.g. facility inspections). On the down side, rotary wing UAVs are less stable than fixed-wing counterparts and also more difficult to control during flight. These platforms maintain directional control by varying blade pitch via a servo-actuated mechanical linkage. Single-rotor and coaxial rotor UAVs are typically radio-controlled, powered by electric motors, although some of the heaviest examples use petrol engines, and they require cyclic or collective pitch control. Multicopters, with an even number of rotors, utilize differential thrust management of the motor units to provide lift and directional control. As a general rule, the more rotors, the higher the payload they can take, and are functional in stronger wind conditions, as the redundant lift capacity provides for increased safety, and more control in the event of a rotor malfunction or failure. A few examples of UAV platforms searching a combined solution have already emerged, combining rotary and fixed-wing technologies (Cetinsoy et al. 2012), providing flying stability and manoeuvrability (e.g. Flying Wing, Songbird 1400).
Sensors on-board UAV for monitoring oil and gas pipelines
The type and quality of sensors carried on-board the flying platform determine the final information obtained from the mission. Although the range of sensors available for small-scale UAVs is forever increasing due to miniaturization and advancements in battery technology (Table 1), limitations associated with size, weight and mechanics still remain (Allen et al. 2014). Selecting a combination of platform and sensor to provide the necessary data in adequate conditions for monitoring and mapping oil and gas pipelines remains a challenge, and for some of the most adequate oil or gas leak detection techniques (e.g. fluorescence), there is still no sensor (e.g. laser fluorosensor) adapted to UAV platforms. The main sensor types with commercial adaptations to UAV mechanics that can be used for monitoring oil and gas pipelines are listed in Table 1.
Table 1 Selection of sensors suitable for monitoring oil and gas pipelines; strengths and weaknesses for the purpose and typical performing tasks
The essential difference between active and passive sensors originates from the source of energy illuminating the target objects (passive sensors rely on the sun, active sensors emit some kind of radiation themselves). This essential difference translates into missions with an active sensor requiring higher lifting and carrying capacity UAV platforms. The capacity to perform a particular task and to work under certain environmental conditions (e.g. topography, weather) is sensor dependent (Table 1). For instance, optical sensors measure radiation in the visible and infrared part of the EM spectrum relying on the sun for illumination; these sensors are suitable in daylight conditions, but even mounted on UAV flying at low altitude, they can be limited by clouds, haze or smoke. Multispectral (MS) sensors measure multiple spectral wavelengths simultaneously providing information that can be visually or automatically interpreted. For a given location, algebraic combinations of values in various spectral wave bands (e.g. vegetation indices) can be useful to detect environmental features, identifying plant stress, disease and nutrient or water status. As a chemical plant stressor, oil spills leave a spectral mark on the plant, that may be identified from the air with MS cameras on-board UAVs. Lightweight thermal still cameras and video have been adapted or specifically developed for use in UAVs (e.g. FLIR Vue); thermal sensors can provide data portraying thermal differences in space and time, revealing the presence of hydrocarbons (i.e. spill, leak) by a strong and increasing contrast in temperature. Video capture and processing enables operational inspections to take place in real-time, without the need to pause production or put ground crew in harm’s way.
The overall range of opportunities provided by optical sensors is constrained by issues concerning the digital frame cameras that can be deployed on lightweight UAVs. The weight and size of the camera, the type of lenses and the spatial resolution, and image footprint size relative to the available UAV payload are limited. Low flying altitudes in small-scale UAVs determine the need for high framing rates and large longitudinal and lateral overlapping of images. Furthermore, basic UAV configurations generate non-metric images, and exposure times are very short to help combat the effects of platform instability due to speed, roll, pitch and yaw.
Active sensors emit some kind of radiation measuring the fraction reflected by the target objects and the difference in time between emission and reception. Active sensors require power supplied by a source, adding weight to the aerial system which makes active equipment less versatile for use on UAVs than passive equipment. Lidar and RADAR sensors have been adapted for use in certain UAV configurations, reducing weight and using specific mechanics (e.g. gimbals). Still, the miniaturization of these active sensors is a remaining challenge. Lidar facilitates generation of very high-resolution (cm scale) surface models and accurate infrastructure 3D models, enabling identification of ground small-scale changes over time before they become a hazard and identification of irregularities in the infrastructure. Likewise, interferometric processing of synthetic aperture radar (SAR) data enables land subsidence mapping to the millimetre scale facilitating the identification of subsidence patterns long before a landslide or other disaster occurs (Bianchini et al. 2013); periodic SAR surveys constitute an alerting system to secure infrastructure like oil or gas pipelines.
For detection of gas (e.g. methane) presence, flux detectors originally designed for hand use (e.g. Laser Methane Copter (LMC), Pergam) are suitable to mount on UAVs. Moreover, other high precision gas detectors for UAVs are under development (Allen et al. 2015). Laser gas detectors beam radiation of a gas-specific wavelength (e.g. 1.65 nm for methane) and measure the backscattered radiation when part of it has been absorbed by the gas cloud. Differential absorption Lidar (DIAL) technology (Zirnig et al. 2005), currently operating on manned helicopters, beams and records light pulses of two different wavelengths—the measurement wavelength is absorbed by the gas (if present) whilst the reference wavelength is not. On-board UAVs, laser gas detectors can detect (Horn 2016) and provide accurate measures of gas leaks (Hodgkinson et al. 2006), although they have an inherent disadvantage of small sampling capacity (Bretschneider and Shetti 2015). Gas (methane) cameras enabling visualization of the gas presence might be useful for identification of pipeline leakages; the adaptation of these sensors to small-scale flying platforms has still to be realized. Not aiming to be exhaustive, Gómez and Green (2015) provide examples of commercial sensors from most of the techniques listed in Table 1 which are adapted to small UAVs.
Auxiliary equipment
To make a UAV mission successful, the aircraft and main on-board equipment (e.g. sensor) are supported by a series of systems and elements. Amongst the most relevant, supporting equipment include systems dedicated to position and navigation, to autonomous flight and for communications. Additionally, the need of elements for the launch, recovery and retrieval, and the mechanics and payloads are UAV and mission dependent. The authors briefly note them here, but for more details, a comprehensive review of commercial auxiliary equipment is compiled in Colomina and Molina (2014).
The positioning and navigation systems play a crucial role in UAV missions, since the location of the UAV has to be known and controlled at all times, be it by the remote operator, or by the autonomous pre-programmed flight planner. The quality of lightweight and compact GNSS equipment now available and capable of receiving signal from multiple satellite systems (e.g. GPS, Galileo, BeiDou) provides for the acquisition of high accuracy location information (Colomina and Molina 2014), especially when operated as differential GPS (DGPS), and facilitates all UAV navigation. For remote control (from the ground, air or sea), where non-autonomous operations are necessary, radar (Jang et al. 2015) or radio (Nitti et al. 2015) tracking solutions are required (Austin 2010). Basically, the radar tracking system fits a transponder to the aircraft that responds to a radar scanner emitting from the control system (CS) and enables the CS to control the aircraft position (bearing and range). With a radio tracking system, a signal carries data informing the CS of the aircraft bearing; the range is determined from the aircraft to CS signal travel time.
Autonomous capacity to both take off and land, as well as to fly along a pre-defined path with n-waypoints pre-programmed by the UAV operator, is desired where the survey is likely to be repeated at certain time intervals ensuring the repeat imagery covers exactly the same area. Likewise, the ability of the platform to navigate amongst obstacles in the flight path and the sense and avoid technologies are advantageous for safety and for systems expected to fly in enclosed spaces. Ideally, the control software (e.g. Mission Planner and APM Planner from Dronecode©) enables pre-programmed autonomous flight as well as manual control when the operator decides to modify the original plan. Existing and economic systems can be easily programmed and monitored with the aid of a smartphone or tablet.
The communication between small UAVs and the CS is usually through radio frequency, commonly in the 900 MHz and 1.2-, 2.4- and 5.8-GHz bands. Uplink transmission (i.e. CS to aircraft) consists of a flight plan, real-time flight-control commands, control commands to the different payloads and updated positional information. The downlink information (i.e. aircraft to CS) consists of the payload data (e.g. imagery), positional data and aircraft housekeeping data (e.g. battery or fuel state). Two different frequencies are necessary to keep the transmission of command information and sensor acquired data independently, avoiding interference. The complexity, weight and cost of the communication equipment are determined by the range of operation possible, the sophistication of the payload transmission and the need for security.
Fixed-wing vehicles require additional equipment to assist with launch and recovery. Launch equipment is typically an acceleration ramp with a trolley to provide the aircraft enough speed to sustain flight; the acceleration is provided by compressed air or by rocket. Recovery equipment can be a parachute installed within the aircraft, combined with a means to absorb the impact energy (e.g. airbag or an easily replaceable piece of material). For most UAV acquiring data, an important requirement is the ability to control the sensor pointing direction and to maintain the sensor orientation during the flight. Furthermore, securing the sensor on the front or underside of the UAV platform is crucial, because rotor vibration and gust instabilities very easily translate into blurred images and shaky videos (Bereska et al. 2013) particularly when flying too fast. Simple solutions for stabilization include damping the sensor (e.g. camera) with a mounting bracket and rubber mounts in between the UAV and the sensor; this will help in reducing or eliminating what is known as the ‘jello’ effect on video imagery. An elaborated solution with specific mechanics adapted to both the platform and sensors is the addition of a gimbal, a precision-engineered component that provides control of pan, tilt and yaw—in case of 3D gimbals.