One of the biggest issues when planning a Beyond Visual Line Of Sight (BVLOS) LOS operation is that the applicable rules and regulations vary markedly between countries ^[1][20]. To this effect, the ability to conduct BVLOS operations and under what conditions are not standardised. 

Primary radar is a technology that was widely used within manned aviation to separate aircraft from one another, and it is still sometimes used today as a back-up to the Secondary Surveillance Radar (SSR) systems, which are dependent upon aircraft transponders, or as a ground radar while aircraft are on the ground ^[2][39]. However, primary radar can also be used on board unmanned aircraft as a means of achieving the DAA requirements associated with flight in non-segregated airspace ^[3][40]. Unlike SSR and the use of Automatic Dependent Surveillance - Broadcast (ADS-B) In and Out, primary radar does not depend on aircraft having transponders on board in order for them to be detected, and it also detects objects other than aircraft, such as buildings and trees, that would otherwise not be picked up ^[4][41]. This is important, as many countries allow the flight of aircraft with no radio on board (known as NORDO aircraft). These NORDO aircraft can only fly in uncontrolled airspace, but unlike other aircraft, they do not make radio calls with position reports, and they are also not listening to position reports of other aircraft. They are entirely dependent on visual detection for avoiding other aircraft. Equally, in many countries, ADS-B transponders are not required to be installed on aircraft for flight in uncontrolled airspace, so they would not be picked up on SSR. These aircraft would be picked up by primary radar, allowing for NORDO aircraft and aircraft without transponders to be picked up. Primary radar is also less effected by sunlight and adverse meteorological conditions, such as fog, dust, rain, snow, or smoke, all of which are a major hindrance to an optical system and which may also affect lidar ^[4][5][41,42]. However, primary radar is not without its drawbacks as a DAA tool; namely, because it has a lower level of locational accuracy (due to longer wavelengths), it is often too large and has too much required power output to be fitted on unmanned aircraft, so it poses challenges for signal processing, and it is very costly to purchase and maintain ^[4][5][6][41,42,43]. However, it is also possible to use a more powerful primary radar as part of the ground control station instead of on the aircraft. This does not avoid the high costs of purchasing radar, however, but it may be a good alternative if the aircraft is not large enough to have onboard radar. If properly deployed, it has been shown to be able to detect cooperative and non-cooperative aircraft, but it is often still combined with other DAA technologies for maximum effect ^[7][8][44,45]. There are also considerations in terms of where this is positioned, given the topography of different forest environments, with the potential requirement to use visual observers outside of the radar’s coverage ^[8][45].

ADS-B is a DAA system that relies upon the unmanned aircraft having a transponder on board that broadcasts its position, registration (or identity if the unmanned aircraft is not registered), direction of travel, and speed over an encrypted data link ^[9][46]. The ADS-B Out function is when the aircraft sends its own data, while ADS-B In will provide information about other ADS-B equipped aircraft ^[10][47]. ADS-B In is commonly used on unmanned aircraft operations, as it is relatively low-cost and can be used with small, unmanned aircraft as an aid to situational awareness ^[11][33]. ADS-B Out is more complicated, as the required power output and potential for interference with other onboard systems mean that only larger unmanned aircraft could have a conventional ADS-B Out system. However, there are lower power alternatives to the ones used on manned aircraft that can be applied to smaller unmanned aircraft, albeit with reduced broadcasting ranges ^[12][13][14][48,49,50]. The suitability of ADS-B systems as a DAA tool depends upon the airspace where the aircraft will be used. If it will be used within controlled airspace, this would bring the unmanned aircraft up to similar standards applied to manned aircraft, and Air Traffic Control (ATC) will be able to ensure separation between the unmanned aircraft and any other aircraft within the operating area. Even in remote areas, having a lower powered system is an additional safety layer that, if a manned aircraft that has a transponder on board was approaching the unmanned aircraft, could trigger some form of alert or warning. However, if the other aircraft does not have the ADS-B transponder, it will not be picked up with this tool alone.

Lidar (acronym for light detection and ranging) is a tool widely used within forestry for remote sensing and data collection, particularly for measuring tree height and collecting inventory data ^[15][16][51,52]. Lidar has a much shorter wavelength compared with primary radar, which means that its effective range is also shorter, though with better range and azimuth accuracy ^[4][41]. However, for low-level flying, lidar is capable of identifying obstacle collision hazards (e.g., trees and vegetation, powerline cables and manmade structures, animals, and so onetc.) and helping to maintain the unmanned aircraft at a certain separation above the ground ^[4][17][41,53]. Lidar’s limited effective range is approximately 50–100 m ^[18][54], so it is not an effective DAA tool against other aircraft. Rather, its application as a DAA is usually in a complementary role by providing low-level detection of obstacles. Nonetheless, recent research has shown that unmanned aircraft can perform autonomous sub-canopy flights using only lidar technology to avoid obstacles along the flight path ^[19][20][55,56]. SLAM-based UAV DAA systems have the dual benefit of being able to capture a detailed three-dimensional model of the forest beneath the canopy, which can be used for stand analysis ^[21][57]. SLAM algorithms also work by matching geometric features from their environment to aid in situational awareness and navigation, so a forest is arguably an ideal situation for their use, as tree stems provide a feature-rich environment ^[22][58]. This shows great potential for future research as lidar is a relatively inexpensive DAA tool if used in this way. The flight is sub-canopy, which means that the airborne risk is zero because no manned aircraft can enter the operating area, offering potentially more accurate data for issues, such as sub-canopy snow depth or sub-canopy invasive species, than current methods that use unmanned aircraft to fly above the canopy ^[23][24][59,60]. Nonetheless, the corollary is that flight at lower altitudes means that less area can be covered than using overflight, so the viability of this BVLOS technique is highly dependent on the size of the forest environments and the quality of data that is required.

With the development of machine learning and deep learning, it is now possible to have on-board DAA systems that utilise Electro-Optical (EO) sensors. EO systems work by sensing light waves and utilise software to analyse the incoming scenes and detect any incoming threats to the aircraft ^[25][61]. While as recently as 2008, the ability of such systems was questionable because they could not detect small aircraft such as microlights ^[26][62], significant progress has been made in recent years to improve their reliability ^[27][28][29][63,64,65] as well as how they can be adopted as part of multi-sensor architectures ^[30][66]. EO systems are typically low weight, small, and are often inexpensive ^[25][61]. However, they do not perform as well in visibility limiting meteorological conditions or in poor lighting ^[25][61]. They also suffer from a limited range, and often, they have a limited field of view, which mean that they are often not a complete solution for DAA ^[31][67]. Anecdotally from projects that the authors have worked on, providing a 360-degree field of view can significantly impact upon weight and the power output required, so this may not currently be a viable solution. Additionally, processing power and resolution for images are also challenges with this technology ^[6][43]. However, EO sensors provide a significant enhancement when used with other DAA tools, particularly when combined with primary radar ^[30][66].

FLARM, short for flight alarm, is a system used in general aviation that informs the pilot of any nearby aircraft (manned or unmanned) or parachutes that have FLARM on board and locates them relative to the FLARM-equipped aircraft’s built-in Global Navigation Satellite System (GNSS) receiver and altimeter ^[32][33][34][68,69,70]. FLARM can be meaningfully compared with ADS-B to consider why it may be chosen in lieu of—or additional to ADS-B—as a means of DAA. FLARM only has a range of up to 5 nm (9.26 km). Therefore, it cannot detect high velocity aircraft with enough time to safely respond, does not integrate with traffic collision avoidance systems, only works in the region it is purchased, and does not provide suggestions to pilots as to how to resolve any collision ^[32][68]. Despite these disadvantages, it also has some advantages compared with ADS-B; namely, it is cheaper, quicker, uses less energy, and can operate independently from ATC ^[32][68]. As noted earlier, forest environments are typically in remote areas, away from controlled airspace or significant aviation activity. The ability of FLARM to operate independently of ATC is, thus, a major advantage, and the disadvantages of range and inability to detect high velocity aircraft are unlikely to pose any major hindrance to the typically low altitude operations that take place in forestry. Thus, FLARM can provide a low-cost addition as part of a combination of DAA measures to help detect other aircraft.

TCAS is widely used on commercial aircraft in combination with ADS-B. Unlike ADS-B, TCAS actively interrogates aircraft transponders to identify intruders and their range, providing both a traffic advisory (i.e., that the traffic is there) and a resolution advisory (i.e., how to respond to avoid a collision) ^[35][36][71,72]. However, TCAS only works alongside ADS-B, may not be able to handle multiple intruders, and smaller unmanned aircraft may not be able to be equipped with TCAS due to payload constraints ^[37][38][73,74]. Nonetheless, if flights are happening where there are commercial aviation operations present, especially if higher altitudes are being used, then TCAS can provide a useful addition. It is highly situation-dependent, but for most forestry operations, TCAS will not provide any additional safeguard; however, it is worth mentioning in case it is relevant for a particular operation based upon the operating environment.

Command and Control (C2) links provide necessary bi-directional communication to and from the unmanned aircraft, normally consisting of commands from the operator’s Ground Control Station (GCS) and telemetry on the unmanned aircraft’s operation back to the GCS. C2, by definition, is a separate data stream to payload data (such as imagery) returning from the aircraft. Depending on the technology implemented on the UA to meet the necessary DAA requirements, it may or may not be necessary to have this payload data returned to the GCS; however, C2 links are required for all UA that are not operating completely autonomously (which is beyond the scope of this paper). While radio communication is not strictly limited to VLOS, radio signals refract and attenuate when they encounter obstacles, such as trees, terrain, or atmospheric factors ^[39][40][22,23]. The amount of degradation experienced varies greatly, depending upon a large number of variables. Best practice is to consider radio transmission to also have a line of sight, even if the UA is not able to be visually seen by the operator due to distance. Operating BVLOS in a forest environment, therefore, introduces a number of challenges.

Many Consumer Off The Shelf (COTS) unmanned aircraft operate within the Industrial Scientific Medical (ISM) bands: primarily on 5.8 and 2.4 GHz. ISM bands are used due to being gazetted under a General Use Radio License (GURL) for Short Range Devices (SRD) in most countries, thereby not requiring any other licensing or certification other than complying with ISM requirements (established by ITU Radio Regulations worldwide, the primary requirement being equipment operating within ISM bands must tolerate interference generated by other ISM applications, and users have no regulatory protection from ISM device operation in these bands.). Due to the relative ease of operating in the ISM bands, many consumer electronics beyond the scope of UA operate on these same bands, such as Wi-Fi and Bluetooth. The prolific use of the 5.8 GHz and 2.4 GHz bands can result in these bands being overpopulated with equipment operating within them, which has led to a number of incidents involving unmanned aircraft ^[41][75]. However, in the comparatively remote locations of forest environments, the flooding of these bands is unlikely. Although the remote locations of many forest operations may result in relatively empty ISM bands, there are other possible radio frequency hazards of remote area operations such as backhaul internet installations and Peer to Peer (P2P) Wi-Fi (for instance to and from milking sheds). Both of these examples use highly directional radiation patterns resulting from using high gain antennas. P2P Wi-Fi use is largely undocumented, and due to the relatively low levels of regulation of the ISM bands, it often operates at power levels beyond legal specification. Backhaul internet and other radio frequency installations may be mapped by commercial service providers ^[42][76]. Mitigations for these installations can include conducting RF surveys of the area using RF spectrum analysers. Using backhaul maps and RF surveys in operational planning, one must also consider possible interference of harmonics; for instance, a 1.2 GHz transmitter will make a 1st harmonic at 2.4 GHz, and a 2nd harmonic at 3.6 GHz.

The relatively high frequencies of the 2.4 and 5.8 GHz bands allow for a higher data rate, providing necessary bandwidth for digital video payload links, as well as the comparatively low bandwidth requirements for C2. Conversely, high frequencies are more easily attenuated by the environment and obstacles. Due to this many modern COTS unmanned aircraft operate on both bands: 5.8 GHz when received signal strength is high, and 2.4 GHz when attenuation happens on 5.8 GHz. This fall back to lower frequencies is associated with a lower data bandwidth and lower bitrate video stream. The amount of radio attenuation experienced by a transmitter varies drastically depending on numerous factors. For forest environments, the density and water content of foliage will have a dramatic effect on the performance of a payload or C2 link. Comparatively lower frequencies (such as the commonly used and gazetted 915 MHz band) are used for C2 links in commercial and military applications due to the lack of radio penetration of higher frequencies. This often results in the operations losing the payload data stream; however, maintaining command and control can allow the aircraft to reposition and resume payload data transmission. These lower frequencies do not support bandwidth capable of conveying data such as video in any meaningful resolution with current technology.

The issue of higher frequencies experiencing greater attenuation and lower frequencies having lower bandwidth is one of the primary issues to solve for BVLOS, particularly in forest environments where the operator and GCS might be below the tree canopy. However, there are a number of available and feasible solutions. Traditional unmanned aircraft C2 links are a Point to Point (PTP) technology, meaning one GCS transmitter/receiver communicating to one aircraft. PTP architecture is comparatively simple, not requiring any further equipment to install (or batteries to charge), and is the most common form of C2 and payload datalink employed in UA operations. Moving beyond PTP architecture can solve many of the inherent problems of radio frequency line of sight, particularly on ISM bands.

Mesh networks are one such radio architecture beyond the PTP links. Unlike PTP networks, mesh networks employ multiple receivers and transmitters, known as nodes, on a network. Nodes are dispersed strategically over the operating area. C2 and Payload data can be transmitted between these nodes over multiple short line of sight paths at higher frequencies, and finally, to and from the aircraft. This offers a number of benefits, such as high bandwidth, as well as the possibility for redundancy, depending on network configuration (data can be broadcast over multiple pathways to get to the aircraft). The downsides of the mesh network are the initial infrastructure cost and the relatively permanent nature of the network (particularly when the mesh is installed over a forestry block, mounted, for instance, in or above the forest canopy). Mesh networks could also offer further DAA options, such as stationary cameras (performing computer-based vision or electro optical traffic tracking) or ADS-B transceivers installed in each node. This concept introduces the possibility of operating relatively unsophisticated UA over a “smart forest”.

Another solution to radio frequency line of sight is the use of a single repeater station which has radio frequency line of sight to the operator’s GCS and the aircraft, while being positioned between them. An example of a repeater would be an unmanned aircraft operating with a hill between the operational area and the operator, with the repeater positioned on top of the hill. This solution offers greater simplicity than a mesh network, and it is likely easier to install. However, it does not offer redundancy. The repeater concept can be taken further, with the repeater mounted on a second unmanned aircraft acting as a satellite and providing line of sight to both the unmanned aircraft and operator. This repeater unmanned aircraft would lend itself to being fixed-wing, lighter-than-air, or a tethered multi-rotor, capable of relatively long loiter times over a relatively fixed position.

Cellular telephone networks, such as 3G/4G/5G have the required bandwidth to handle C2 and payload data streams, and they have been used to operate unmanned aircraft, offering RF ranges as far as the existing cellular network extends. Benefits of this architecture is the use of existing infrastructure rather than an ad hoc network. Cellular networks operate on multiple frequencies and can have exceptional high data throughput capabilities. However, cellular networks are rarely installed with particularly robust coverage in the remote locations where forestry exists. Cellular networks also introduce a large degree of latency in the link, increasing the time between operator input and the UA responding or payload data returning to the GCS. This latency may affect DAA effectiveness and response times.

Other solutions may employ using segregated bands with radio frequencies licensed and reserved for the operator (often at higher than ISM output power and, therefore, range). This, however, incurs regulator and hardware costs. Satellite links are also an option; however, these are more aimed at large, medium/high altitude long endurance (MALE/HALE) unmanned aircraft and military applications, due to their high costs and difficulties with fitting antennae to smaller aircraft ^[43][44][77,78]. These may be worth exploring, due to their benefits, but only when there is a sufficient budget or if using large equipment.

One question to ask before performing a BVLOS operation is whether it would be easier to vary typical Extended Visual Line of Sight (EVLOS) requirements instead. Standard EVLOS requirements involve either the pilot or visual observers having unaided visual line of sight of the aircraft such that they cannot use any tools, such as binoculars, to enhance their vision ^[1][20]. However, sometimes regulatory authorities approve operations that vary this requirement and where visual observers use technology such as binoculars, telescopes, night vision goggles, or zoom lenses to observe the aircraft and airspace at a much greater distance than what is achievable using human vision alone. There needs to be some DAA capabilities built into the aircraft in order to justify an EVLOS+ operation because the use of binoculars limits the field of view, and the magnification means that smaller tracts of airspace are being observed at any given time ^[45][79]. However, the sophistication of the DAA technology can be significantly less than that for a BVLOS operation because the usual VFR rules of see, detect, and avoid can largely be achieved in the usual manner (i.e., visual scanning) under EVLOS+. It is questionable whether EVLOS+ is safer than a BVLOS operation, however, because it works with existing rules and regulations, and it may be easier to achieve the requisite standards for such an approval. OThe authors could only find one mention of a similar operation ^[46]in the coliteratuld be foundre [80], indicating that this could be an interesting area for future research. Industry experience suggests that this can be more cost-effective than BVLOS (by avoiding the need for highly sophisticated DAA technology) and more practicable than EVLOS (requiring multitudes of visual observers if in a large operating area).

A simple means of removing airborne risks associated with BVLOS flight is to fly in airspace that does not contain other aircraft ^[1][20]. Most commonly, this would be Restricted Airspace (RA) but could be other segregated airspace, such as Military Operating Areas (MOA). As noted earlier, forestry companies can apply to have the airspace above their estate (up to a certain level) made into Restricted Airspace (either temporary or permanent in nature) such that they can control whether other aircraft can enter that airspace or not ^[1][20]. Countries do vary on their likelihood of giving this airspace and how easy it is to apply for it. However, the general consideration is whether the imposition of Restricted Airspace would disrupt extant aviation operations ^[1][20]. Given that forested areas are typically remote, this approach may make sense for larger estates to avoid the need to purchase sophisticated DAA tools for the aircraft.

In manned aviation, air band radio is considered as the second most important aid to situational awareness, sitting only behind visual scanning ^[47][81]. There are two considerations for using air band radio with a BVLOS operation. The first is listening to broadcasts on the local frequency. This is a must for a BVLOS operation that is being conducted in non-segregated airspace. This is because listening to position reports will allow the pilot to form a mental model of how traffic is acting in the airspace surrounding them, thus providing useful information to help detect and avoid other aircraft ^[48][82]. A recent study showed that just over a quarter of unmanned aircraft users listen in to air band radio when conducting operations, primarily when operating above 400 ft AGL, within 4 km of an aerodrome, or within controlled airspace ^[49][21]. These correspond with the types of airspace where this is most useful because there is the greatest potential for manned aircraft to also be operating within these areas ^[1][20]. To be able to make sense of radio calls, there will need to be some training, as well as an awareness of where the operation is taking place so that the pilot will be aware of which locations’ position reports will be made in relation to them (i.e., 5 nautical miles southwest of “location”). For BVLOS operations, it is also advisable to make position reports for the unmanned aircraft to aid another pilot’s situational awareness. However, to be able to make radio calls requires a Flight Radio Telephony Operator’s (FRTO) rating, which is typically held alongside a pilot’s license ^[1][20]. Some countries, such as Australia, have made this easier by providing a means for unmanned aircraft pilots to have an FRTO rating independently of a pilot’s license ^[50][83]. If the operation is taking place in a country where it is not possible for unmanned aircraft pilots to make radio calls, and the pilot in command for the operation does not have a pilot’s license, it may be necessary to have someone with a pilot’s license in the support crew. This relates back to meeting the requirements of the airspace that is being flown within, which, outside of segregated or controlled airspace, means making position reports so that other pilots can maintain situational awareness of traffic in the area.

JSAs are a formalised way of showing that steps have been taken to reduce risk levels to people, property, and other aircraft to as low as reasonably practicable ^[49][21]. The great thing about them is their flexibility, whereby hazards and associated risks pertinent to a specific operating area can be taken into account and managed, in combination with other operational procedures, to manage more general risks. For example, many airports have their own requirements for flying in their airspace, which are additional to any local aviation regulations ^[51][84]. Just over 20% of unmanned aircraft users typically conduct a job safety assessment prior to flight ^[49][21]. Being familiar with the operating area, aware of risks, and planning flights accordingly is critical for effective risk mitigation ^[52][85], and JSAs provide evidence that this has taken place.

One of the most important aspects of pre-flight planning for a BVLOS operation is to ensure that all aeronautical information is taken into account. At the most basic level, this involves checking what airspace the operation will be within, the requirements of that airspace, and how to comply with those requirements (or if it is even possible to comply with those requirements). Some of the key sources of aeronautical information for unmanned aircraft operators include the Aeronautical Information Publication (AIP), AIP Supplements, Visual Navigation Charts (VNCs), and Notices to Airmen (NOTAMs) ^[1][20]. Each one of these provides information that is relevant for different time intervals. AIPs provide aeronautical information that is of a lasting nature and is essential for air navigation, while the AIP Supplements provide temporary changes to information published in the AIP ^[1][20]. VNCs are probably the most used of the sources of aeronautical information because these visually depict this information, including airspace boundaries, aerodrome locations, special use airspace, and other useful information ^[49][53][54][21,30,86]. Reading VNCs has been established as a key risk mitigation because of how it assists with flight planning and situational awareness of the surrounding airspace ^[52][85]. Another source of aeronautical information, that unmanned aircraft operators should check prior to operation, are NOTAMs. Unlike other sources of aeronautical information, NOTAMs are issued daily and contain information about temporary changes to aeronautical information or procedures, as well as any temporary hazards that should be taken into account when planning a flight ^[49][55][21,87]. This has also been identified as a key risk mitigation step that unmanned aircraft operators should take to help mitigate airborne risks ^[56][36]. The issuance of NOTAMs may also a useful safety precaution for BVLOS operations, as this will make manned aircraft pilots aware of the ongoing operation, the operating area, and times of operation. The importance of this would depend on what airspace and what altitude the operations take place within. When operations are taking place in airspace that is shared with manned aircraft, then, it would be advisable to issue a NOTAM.

The least predictable component of any system is the human operator ^[57][88]. One way in which this can be more standardised is through formal training and competency requirements that form part of the overall operation. There may be different sources of training and competency requirements. Some requirements are stipulated by government regulators and form part of an application for a license or certification to carry out a certain activity. However, companies often also have internal training and competency requirements that form part of their safety assurance. When considering a BVLOS operation within a forest environment, there are some unique considerations for training. Firstly, the operating environment presents a number of difficulties that have already been outlined in this paper. Pilots and support personnel need to be familiar with these peculiarities and how to manage them. Secondly, there are some unique hazards within this environment that require thorough training on emergency procedures. One salient example is the risk of creating a forest fire due to an accident. Lithium polymer batteries and petroleum (the two most common energy sources) are inflammable and highly flammable, respectively. The organisation should have considered how it would manage post-crash activities when writing their operating manual; however, staff require training on these emergency procedures to ensure that these can be carried out in a timely and coordinated manner. Thirdly, the highly automated nature of the systems means that the role of a pilot is more one of monitoring systems rather than conventional “hand on stick” flying. This means that they need a thorough understanding of the systems that are being used, their failure modes, and when intervention may be necessary or desirable in the interests of safety. Because there are significant differences between platforms and ground control systems, this training needs to be bespoke for the particular operator. Lastly, an assessment of competence cannot be restricted only to flying ability; competency needs to be assessed more holistically. The term operational competency tends to be used for more novel applications, such as BVLOS in forestry, whereby pilots and support crew need to demonstrate competency in the overall operation, including pre-flight procedures, in-flight procedures, post-flight procedures, emergency procedures, and so on. This often means working with training providers to create bespoke competency assessments that provide a sufficient level of safety assurance for the human element of the operation.

BVLOS operations take place over larger operational areas, where the ability to perform BVLOS is advantageous, which also means that operations are typically longer endurance than conventional operations. In such instances, it is important to consider duty limits for pilots and visual observers to avoid the negative consequences of fatigue. Fatigue is a state of deep tiredness that impairs performance, particularly motor skills, situational awareness, vision, and personal standards ^[57][88]. There are a number of ways to manage fatigue. Primarily, these centre on schedule optimisation, rest breaks and naps, fatigue aids (e.g., caffeine), and ensuring that crew are practicing good sleep hygiene ^[58][89]. The schedule optimisation is aimed at avoiding duty periods that are excessive and ensuring sufficient rest periods between duties. Rules and guidelines vary internationally for manned aviation. However, differences exist based upon whether the operation is a single crew operation or multi-crew operation and whether the flight is done under Visual Flight Rules (VFR) or Instrument Flight Rules (IFR). To generalise those requirements, a single duty period should not exceed 8–11 h, a pilot should not be scheduled for more than 30–35 h in 7 consecutive days, 90–100 h in 28 consecutive days, or 250–300 h in 90 consecutive days ^[59][90]. In addition to this, employees need to ensure they get at least 8 h of sleep per night, plan sleep ahead of time, and report fatigue ^[60][91]. During multi-crew operations, it may also be possible for an individual crew member to take a short break, which sees some short-term improvement in alertness, or even to take a nap (10–30 min), allowing 30 min post-nap to prevent sleep inertia ^[58][89]. Caffeine is also a useful fatigue countermeasure that increases alertness. However, it does take about 30–50 min to act, its effects wear off after about 4–5 h, and high caffeine intake can have an adverse effect ^[58][89]. Personal sleep practices are also critical, with crew needing to ensure that they get enough rest (most people need between 6–10 h sleep), that they try to avoid sleeping at unusual times (sleep at night to align with circadian rhythms), ensure that their sleeping environment is dark and cool, keep their sleep environments as safe sleep zones (e.g., never use the same area for sleep and work), and avoid detrimental sleep aids such as alcohol, among other things ^[61][92]. As can be seen thus far, fatigue management is both a crew responsibility and an organisational responsibility. If conducting BVLOS operations, which tend to be more complex, it is critical that flight crew do not have their performance impaired due to fatigue. Despite the importance of this topic area within manned aviation, research into pilot fatigue for long-endurance unmanned aircraft has primarily been within military applications ^[62][63][64][93,94,95] and often calls for further research due to the lack of data on this issue. This presents a potentially useful area for future research, given the likely increase in long-range BVLOS operations in coming years.

While not necessarily the case, most BVLOS operations require multiple crew members. In such instances, it is important that all crew members are empowered to perform their roles effectively. This process is termed Crew Resource Management (CRM) within aviation and is a compulsory subject area for the issuance of pilot’s licenses. Good CRM means that the tasks associated with a particular flight are organised and distributed such that all of the crew’s resources are applied in a safe and efficient manner ^[57][88]. For this reason, it is important that organisations doing BVLOS operations in forest environments have ensured that CRM training is provided, either internally or externally, to staff to ensure that multi-crew operations are effectively managed. While limited, there is already some literature that supports the need for CRM training for large and complex unmanned aircraft operations ^[65][66][67][96,97,98].

An area that has only received limited attention is the design of the ground control systems in such a way that takes into account the human body and expectations around how things should be controlled. Anecdotally, the authors have come across systems that cause discomfort or are confusing for a human operator were came across. There are no international standards enforced upon the design of ground control systems, however, and that is true for many aspects of unmanned aviation. Some have suggested that standards could be borrowed from closely related areas, such as general-purpose computer workstations ^[68][99]. However, others have suggested that a combination of human–computer interface standards and conventional cockpit design is required as ground control systems for unmanned aircraft resemble both systems, though they also have distinct differences ^[69][100]. Either way, as BVLOS operations become more prevalent, it is critical that the principles of ergonomics are also applied to the design of ground control systems in order to minimise the potential for design-induced errors.

Drugs and alcohol are major concerns within the aviation context, with states often imposing strict regulations to manage associated risks. Alcohol is widely consumed, but it does cause performance detriments, primarily, in relation to cognitive and sensory functions ^[70][101]. While regulations vary between countries, they usually consist of a minimum stand-down period following any consumption of alcohol called the “bottle to throttle” rule (usually 8 or 12 h), which is the requirement to not pilot an aircraft unless your blood alcohol level is zero and to not pilot an aircraft under the influence of alcohol even if your blood alcohol level is zero ^[71][72][102,103]. This latter point is important in capturing the ongoing impairment associated with hangovers. These are standards that should also be applied to unmanned aircraft operators to prevent any alcohol-associated impairment. Past research also suggests that pilots with previous convictions for driving while intoxicated are twice as likely to be involved in a pilot-error accident ^[73][104], providing a good basis for background checks regarding past compliance with transport regulations.

While seemingly not addressed in the literature (to the best of the authors’ knowledge), one risk that needs to be thought through with BVLOS operations in forest environments is the potential for deliberate interference with the unmanned aircraft being operated. Anecdotally, sthe authors are aware of such instances are within New Zealand. The potential for this is that criminal activity in cultivating illegal drugs such as marijuana and cocaine often takes place in remote forest environments on public lands ^[74][75][76][108,109,110]. As unmanned aircraft become more commonly used by enforcement powers and government authorities, it is possible that an unmanned aircraft being used for legitimate purposes in a forest environment may be confused with law enforcement unmanned aircraft. In such instances, it is possible that the unmanned aircraft may be shot down. This is a risk that needs to be considered when operating over forest environments where it is known that there may be cultivation of illegal crops. Options include not conducting the operation, publicising the operation and its purpose in the hope of getting the word out, or even direct contact with organisations that are known to be involved with such activities. The latter option does also present ethical issues. However, those are outside of scope for this paper.e As this risk appears to be undocumented in the literature, it would be useful to see research on the frequency of illegal interference with unmanned aircraft operations and also what measures operators are using the mitigate the likelihood of such occurrences.