A UK initiative is working on ways to allow UAVs to share airspace with manned aircraft – potentially opening a new market in peaceful uses for unmanned aviation. TIM ROBINSON reports on progress with the ambitious ASTRAEA project.
UAVs, it seems, are never far from the news these days. But usually the media headlines concern ‘killer drones’ – referring to the vast growth in military unmanned air systems (UAS) (to give them their full designation) in the past decade. These UAS have brought unique capabilities to the battlefield, allowing persistent surveillance and more precise and accurate lethal force to be applied. Yet the civilian applications of UAS, though promising, have lagged due to the challenges of mixing unmanned air vehicles and manned aircraft in the same airspace. Once unlocked, some predict that civil UAS could be the airborne equivalent of the mobile phone — a ‘disruptive’ technology whose killer ‘app’ or ‘must-have’ service hasn’t even been thought of yet.
In the UK, research into the this issue is being spearheaded by ASTRAEA (Autonomous Systems Technology Related Airborne Evaluation & Assessment) in a £62m jointly funded industry/government initiative to explore key technologies and operating procedures for UAS to operate in civil airspace without specialised procedures. Its participants include backing from the Technology Strategy Board (TSB) as well as half the funding from UK industry in the form of AOS, BAE Systems, Cobham, EADS Cassidian, QinetiQ, Rolls-Royce and Thales UK. It also has input from other stakeholders including CAA, and MoD as well as universities and other SMEs.
Phase 1 – which ran from 2006-2008, saw technology developed to TRL (technology readiness level 3) and a synthetic environment (simulation) of how a UAS mission might look in the future.
In the current Phase 2 of the project, which runs until March 2013, synthetic environment trials have been joined by live flight demonstrations, with a BAE Systems Jetstream 31 test aircraft acting as a ‘surrogate’ UAV to test control systems. Controlled through its autopilot and instrumented, it is allowing ASTRAEA to demonstrate the ‘virtual certification’ of a UAV — with the advantage of two human safety pilots onboard. According to Dr Darren Ansell, BAE Systems and ASTRAEA technical manager, test flights will begin again in the first two weeks of October. In 2012 another 20 flights are planned between February and mid-summer. Additionally some of the testing will now involve an ‘intruder’ aircraft (a Piper Seneca) to help test the surrogate UAV ‘sense and avoid’ systems.
By completing these dry runs in the form of ‘virtual certification’ it will reduce risk for manufacturers, operators and regulators in the future. However, ASTRAEA points out that the ‘virtual certification’ is not a definite roadmap to acceptance — merely a potential solution. Phase 2 of the programme will also see the TRL raised to TRL 6.
This effort in the UK is not alone and ties in with regulators at both European and international level to understand and prepare for the requirements. However, ASTRAEA is unique in a sense that it is a multidisciplinary effort, looking at technology, procedures and, crucially, regulation as a holistic whole. one way in which UAS integration at a higher level will be helped is in the introduction of NextGen and SESAR future ATM systems in the US and Europe. In these separation control will shift subtly over time into the cockpit, leading to automatic separation control. These 4D trajectory air traffic systems, with airliners automatically sharing information between them to keep a safe distance part, should make it easier for UAS to enter manned airspace seamlessly.
Below is an exclusive interview with Programme Director Lambert Dopping-Hepenstal.
Research at ASTRAEA, though, is not just about building a safe air vehicle – but also about how UAS fit in with the airspace as a whole in a systems-of-systems approach, covering automation, sense & avoid, communications, human factors and ground operations. Says Lambert Dopping-Hepenstal, of the systems approach: “We are looking at what is different from manned systems and what we need to do to understand it.” At the conference, speaker Prof Brian Collins from UCL notes, this complex and multifaceted challenge is a ‘wicked problem’ – not teen slang but a phrase describing a difficult “problem that is not understood until after the solution is found.” And the solution not only involves airspace regulators and stakeholders but also, intra-regulatory co-operation.
A regulators perspective
At the conference, Grp Capt John Clark from the UK CAA provided the regulators perspective. He noted that UAS must work within the current manned airspace rules – with the keywords being ‘equivalence’ and ‘transparency’. A UAS still has a ‘pilot in command’ with ultimate responsibility – they are just separated from the vehicle. Clark warned that it is for industry and the community to prove that it will meet standards – “whatever you propose it must be safe.” The issue at the moment, he explained, is that while segregation of manned and unmanned airspace (with temporary danger areas) ensures no mixing it, limits airspace and is restrictive. However the CAA’s approach, in common with other national and international regulators, would be a step by step one – “safety is paramount”.
ASTRAEA is also working on the question of autonomy or artificial intelligence (AI). Contrary to popular belief (and science fiction) there are no autonomous drones. Truly autonomous aircraft says Nigel Mills, QinetiQ, are a long way away. Instead what we will have are remotely piloted aircraft with autonomous phases of flight, ASTRAEA’s approach, however, goes one further to variable levels of autonomy so the pilot/operator is working in partnership with the AI. The level of automony would vary depending on the task, (take-off, cruising, avoiding action) and whether there were any special conditions such as a lost link and/or a collision course with reduced reaction time. Should the system find itself in this situation of having calculated a collision course and, receiving no instructions (perhaps because of a broken communication link), it will thus automatically take evasive action to avoid an accident. Mills observes that while the human in the cockpit does have advantages (ie working out a more creative solution to a problem by bending the rules) autonomous systems also have distinct benefits. They don’t get bored, don’t get tired, don’t take risks and they objectively re-evaluate new situations. Furthermore, he argues, in the future autonomy may be essential to keep costs down by allowing a single pilot to control multiple air vehicles.
Perhaps the biggest challenge currently holding back the full exploitation of civil UAS is in sense and avoid – the ‘holy grail’ of integration. How does one replace the Mark 1 eyeball that sweeps the sky recognises a tiny speck that may be collision threat, and allows the pilot time to take avoiding action? An easy solution would be to ensure that any UAS operates in controlled airspace, with a filed flight plan, TCAS/ADS-B and under close control from the ground. Yet it is in uncontrolled (Class G) airspace, close to the ground in VFR conditions where the majority of UAS will likely be operating and in which they might encounter other airspace users such as light aircraft, helicopters, gliders or balloons. The range of proposed solutions, electro-optic, radar, ADS-B is varied – as is whether these should be mounted onboard or offboard. Currently the community is leaning towards an onboard electro-optic solution – with a lookout provided by dedicated sensors, but ADS-B and TCAS-compatibility will also have their place. As one speaker observed of the on/offboard sense and avoid question “it will be both – the clever bit will be in fusing the two.” In one respect, however, an unmanned aircraft on potential collision course may have an additional advantage over a manned aircraft – as without crew or passengers, it may be able to manoeuvre harder and more dynamically.
However, some noted that there will be some time before the technical advances being developed by ASTRAEA will filter down to smaller UAS. Philip Church, consultant at Helios notes: “ASTRAEA is developing technical solutions and capabilities that are currently better suited to large UAS. However, the UAS civil market has a demand for the smaller (less than 150kg vehicles) which will also be predominantly flying in uncontrolled airspace mixing with other air traffic. I suspect that sense-and-avoid capability suitable for this market is still some way off.” He also noted the current certification objective of proving ‘equivalence’ will result in an interesting juxtaposition. “For example, if you consider ‘wake vortex avoidance’ as part of the sense and avoid requirement — how do you prove ‘equivalence’ for something a human pilot can’t currently do – or can do but not very well.”
In human factors, too, ASTRAEA’s research is fundamentally assessing the operators/pilots ‘trust’ in the system. Can I trust the computer system to follow my instructions? Chris Noonan, Thales, poses the question: how far will the piloting task migrate to supervision? Because of the time delay (especially in beyond line of sight) he likens the future UAV piloting task to more like a supertanker captain – requiring the pilot/operator to think 10-20 seconds ahead of where the air vehicle will actually be. But ASTRAEA is not only looking at the pilot interface (one solution presented uses multiple iPad/tablet style display screens) but also other factors such as crew resource management, boredom and pilot qualifications. In the last instance, does the pilot need to be certificated on the type of air vehicle or type of control station – or both?
In the military world, high-bandwidth but costly satellite links give operators real-time full screen video for surveillance and targeting. For civil UAS this dedicated, expensive satellite communications solution is not an option – except for in the case of lost link. So civil UAS communication systems will have to reshape and reconfigure itself dynamically as air vehicles automatically and seamlessly switch from ground station to ground station. Beyond line-of-sight they will also have to search and connect to satellites. Essentially it is akin to creating a mobile phone network in the sky so that, to the users, the communications are seamless and vehicles are never out of contact. Indeed, further in the future it may be possible to use the air vehicles themselves as nodes in the network.
But EADS Cassidian, who is working on this part of the ASTRAEA, is going one further. Not only does it see UAVs dynamically switching from station to station, constantly switching to the strongest link, but also it is testing an adaptable intelligent imagery downlink. This link will adapt to lower bandwidth by automatically decreasing video resolution (or even stills) – so that, for example, a coastguard UAS, going to the very limit of a ground stations range, will still be able to return useful data.
However, one area where UAS communications may actually be simpler than in manned aviation is in communications between the UAS pilot and ATC – which could be a direct landline.
So should ASTRAEA and other projects like it succeed, what sort of civil market exists for them? The CAA says there are already 75 small UAS operators flying in the UK but these are under current line-of-sight rules (500m) which limits their capabilities. Is there a market for larger vehicles? At the ASTRAEA conference delegates heard from two potential users – the British Antarctic Survey (BAS) and the Countryside Council for Wales.
The BAS, which operates a small fleet of utility aircraft in the Antarctic (a Dash 7 and four Twin Otters), has already experimented with electrically-powered UAVs for measuring sea-ice. Fitted with sensors its aircraft also carry out geophysical surveys, measure ice depth with radar and magnetic research.
It sees great potential for further use of UAS in the empty skies of the South Pole. In particular lower-cost, multipurpose UAVs could carry out research and scientific missions all year round through the Antarctic winter, when manned aviation is grounded. Because of the remoteness of the bases and the need for all supplies to be brought in, the BAS also sees a role in cargo UAS delivering food, fuel and equipment to scientists in the polar region.
Meanwhile, another speaker, Alan Brown, remote sensing manager from the Countryside Council for Wales, outlined how his agency could benefit from low-cost, regularly updated remote sensing that UAS could provide. It has a need for recently updated aerial photography to monitor wildlife habitats, species monitoring, crops health and forest boundaries. The estimated market for this could be vast, with much satellite imagery judged being too costly and having long revisit times for such local and regional requirements. Brown also notes that a critical niche role, currently unfulfilled, that might be provided by small tactical UAVS is for airborne calibration/measurement of satellite or HALE imagery. For environmental agencies and climate scientists MALE UAS have also another advantage in that using less fuel that manned aircraft they are the ‘greener’ solution. Indeed, a solar-powered HALE UAS, such as QinetiQ’s Zephyr, would be the ultimate in an eco-friendly UAV.
Disaster relief and UAS
As well as these jobs, there is also much interest from law enforcement, fire service, border patrol, scientific research and disaster monitoring communities to name but a few. Indeed, UAS are already one of the first tools when disaster strikes. Predator UAVs were used in the hunt for survivors after Hurricane Katrina hit New Orleans. Meanwhile, the US Air Force’s Northrop Grumman Global Hawk was quickly deployed to survey damage and assess the infrastructure needed for humanitarian relief in the wake of the earthquake in Haiti in 2010.
The need for UAS to go where it was judged too dangerous for humans was even more critical this year – when both Global Hawk and Honeywell T-Hawk UAVs surveyed the damage to the Fukushima Daiichi nuclear plant after the Tsunami in Japan. The smaller VTOL T-Hawk, fitted with radiation sensors, was able to provide crucial information for those assessing the scale of disaster. The bottom line is - UAVs are already saving lives.
Other business cases should not be too hard to find – As Prof Collins noted at the conference, after the first 30mins a traffic jam on a UK motorway costs £1m an hour – a potential business case for a Highways Agency traffic spotter UAV? And while some UAVs are cheap, affordability of more complex systems could be addressed by multi-agency sharing or perhaps by joint procurement.
In addition, the development of a civil UAS industry would also benefit, and may even provide a revenue stream, for small regional airports or general aviation airfields. UAS would be unlikely to be welcomed (or be able to fit in) to Heathrow’s traffic pattern but smaller airports, perhaps near the coast, could boost their revenue by hosting UAS services.
It is also noteworthy that not all civil applications (for example agricultural remote sensing) need instant live video data sent – which will help reduce costs.
These roles, though, are unmanned equivalents of either manned aircraft or satellites. But there could also be even more exciting tasks for civil UAVs in the future. Imagine a personal mini-cargo UAV that could collect your weekly shopping from an automated warehouse? Or perhaps fleets of electric delivery UAVs to deliver high-cost, low volume items, reducing road traffic and cutting emissions? Or perhaps a pay-per-hour surveillance/imagery HALE UAV that users could log into, via a web interface, control the cameras for say pipeline inspection, and then log-out to have the imagery or data sent to them? Sensors, too, could be adapted for the civil sphere – for instance a thermal imager could be, instead of programmed to lock on to T-72 tanks, perhaps be adapted to follow the migration of whales. Already there are minds thinking about new, innovative ways to exploit UAS.
Matternet, for example, is a revolutionary start-up idea that seeks to provide roadless transport for poor in the developing world – delivering medicine, supplies and food directly by a fleet of UAV quadcopters. Using UAVs to deliver aid directly to the people who need it would eliminate the theft of humanitarian aid and reach those cut off through inadequate surface infrastructure. Far-fetched? – perhaps. But it has to be remembered that in less than 15 years mobile phones have switched from a tool of voice communication to an all-purpose personal device capable of playing audio and video media, electronic book, GPS receiver and web browsing. The potential here, then, to alter lives for the better could be amazing.
Manned aviation benefits
Ironically, too, some of ASTRAEA’s most critical breakthroughs in automation and human factors may even spin-off into manned aviation – enhancing safety for everybody. For instance, research into automation may lead to a ‘get-me-home’ or ‘land immediately’ button, that could help prevent general aviation accidents if, for example, the pilot fell ill. It may even prevent a repeat of the Helios Airways 737 accident where hypoxia overcame the crew and passengers, despite the brave efforts of one cabin crew member to land the aircraft.
Human factors investigations into ‘trust’ into sophisticated computer autopilot systems that control UAS may also have important spin-offs into making air transport safer. Since, although modern airliner glass cockpits contain ever more information and are easier to use – there still remains the problem of misinterpreting what the computer is saying, or what the aircraft is actually doing. Investigation into the Airbus A330 AF447 accident in June 2009 for example, revealed the crew had missed that the aircraft had gone into another flight control law mode – removing a critical stall protection function. Reducing these sorts of cognitive disconnects between an organic pilot and a silicon pilot’s brain is thus likely to have safety benefits for aviation as a whole.
An image problem?
Yet, outside the technical challenges, the use of civil UAS has as its biggest cahllenge that of public perception. As noted earlier (almost) all military usage so far has skewed public perceptions that they are ‘killer’, ‘spy’ ‘robots’, evoking thoughts of big brother ‘drones’ snooping far above.
The terminology used also can deceive – UAV (as in unmanned) refers to the vehicle itself – but there is no UAS that is completely free from human control – the pilot is simply offboard (hence RPAV (remotely piloted air vehicle). Yet this new designation might also bring further misunderstandings as the rise of air vehicles with (limited or partial) autonomy in some phases of flight will really mean that they will be ‘remotely supervised air vehicles’. Should RSAV be the next acronym?
However, these cultural concerns over ‘machines taking over’ – which stretch all the way back to Luddites, the Industrial revolution, through to Metropolis, Terminator films and Blade Runner – should not be easily dismissed. The public, politicians and media need engaging and reassuring that these drones can be a ‘force for good’. As a study by aviation consultancy Helios notes: “A concerted effort needs to be made to sell the efficiency, environmental and agility benefits that UAS offer over manned aircraft operations.”
Engaging the public
But this is changing and the sector, and ASTRAEA is now reaching out. A video competition by ASTRAEA, launched earlier this summer, challenged creative, and digital media industries to produce a short film to improve the public perception of UAS. This contest, won by ACC Media of the University of Central Lancashire with the video below seen below.
It usefully highlights to the general public how other everyday objects would be, if viewed through the same prism as UAS, such as automatic doors, laptops or even kettles. An electric kettle with an automatic cut-off might be called an ‘unmanned boiling device’ (UMD) yet provokes no strong reactions that it might ‘deliberately’ go rogue. This video then is an important first step to engage those outside specialist aviation and UAS communities that automation is already a part of their lives, and that designed correctly a civil UAS can be as safe as manned aircraft. (Indeed, it is worth noting that a large proportion of current UAS accident statistics come from human error in the landing or take-off phase. A study in the US found that 71% of Predator accidents between 2003-2006 were traceable to human operator error. This led directly to the US Army specifying an automatic take-off and landing capability for its version of the Predator, the Grey Eagle).
But it is important to remember that the UAS community is not alone in facing these problems. Even in Japan, where new technology is embraced much more whole-heartedly, the growing use (and need) for automation to look after the elderly causes concerns.
It is clear that there is massive potential for civil UAS in the future – opening up new businesses. Yet the challenges remain tough and progress must be slow into this new area of aviation. The air transport system and airspace regulation that has grown up over the past 100 years has been paid for in blood of pilots, aircrew and passengers. But today the levels of risk that were present in the early days of aviation will not be tolerated. UAS have to enter manned airspace safely or not at all. The way forward then will be in a considered, step-by-step measured approach, to satisfy both regulators and the public that UAS and manned aviation can mix.
The civil UAS sector also represents a massive economic opportunity for manufacturers, service providers and pilots – and for those who can crack this issue. As Prof Brian Collins notes of ASTRAEA’s work: “It has huge potential for this country and as export potential.”
The groundwork laid by ASTRAEA today may one day be seen as a critical step in unlocking a massive new sector – allowing civil UAS to save lives, conduct environmental research to help protect the planet and perhaps even deliver food to remote places in the aftermath of disasters. Key breakthroughs here may also reduce accidents in manned aviation. Ultra efficient civil UAS systems may end up not only cheaper but, more crucially, greener with reduced noise and air pollution. Today, ‘drones’ may be known as ‘spies’ or ‘killers’ – but one killer ‘app’ may turn these weapons into ‘man’s best aerial friend’.
Want to know more about this? The Royal Aeronautical Society will be running a two-day Unmanned Air Systems conference looking at the Operational and Technological Readiness to the Civil UAS market in October. For more details click here.