Reliable 'see-and-avoid' technology is critical for safe operation of UAVs in civil aerospace
Ramon Lopez/WASHINGTON DC
As civil aviation authorities wrestle with thorny policy issues regarding certification of unmanned air vehicles (UAV) and their operation in civil airspace, government and industry researchers are keeping the prospects alive for widespread use of civil UAVs.
Because of past military research, initial work is already under way towards developing "see-and-avoid" technologies for UAVs that would allow both manned and unmanned aircraft to peacefully co-exist both over a battlefield and in national airspace.
In the USA, the armed forces must currently seek waivers from the US Federal Aviation Administration at least 60 days in advance of a UAV flight in civil-controlled airspace. Even then, flights are limited to short duration. The existing approval process is cumbersome but necessary until UAV technology is available which offers a level of safety equivalent to that of manned aircraft. Systems are required that provide a see-and-avoid capability for an unmanned aircraft similar to that available to a human pilot operating under visual flight rules.
The American Institute of Aeronautics and Astronautics recently hosted a workshop outside Washington DC during which Lockheed Martin's Rowena Eberhardt provided a status report on joint development of an automatic air collision avoidance system (ACAS) by the US company, Sweden's FMV defence materiel administration and the USAir Force Research Laboratory (AFRL). The auto-ACAS could one day prevent mid-air collisions between vehicles by tracking their direction and trajectories.
Automation is achieved by providing data from each to the digital flight control computer in the other and commanding escape manoeuvres for both the manned aircraft and the UAV. The system only activates at the last possible moment to avoid interfering with normal operations. It performs much like the traffic alert and collision avoidance system (TCAS), which already provides a safety net for manned aircraft, but TCAS needs two pilots to work. TCAS beacons are planned for installation on Northrop Grumman RQ-4A Global Hawk high-altitude endurance (HAE) UAVs to broadcast their positions while airborne.
Several years ago, the AFRL developed an automatic ground collision avoidance system (GCAS) involving a digital database of terrain surrounding an aircraft. The aircraft location determined by radar altimeter and inertial navigation system is continuously compared with the database. An aircraft response model was developed to predict the future trajectory of the aircraft, and a minimum descent altitude (MDA) is placed over the digital terrain. If the future trajectory intersects the MDA, a warning is issued. Lack of action by the pilot produces an automatic climbing turn.
Design was completed by 1992 and the GCAS successfully underwent low-level night attack flight tests on a Lockheed Martin F-16. Four years later, the Swedish air force joined AFRL to expand the limited envelope of the automatic GCAS, which has yet to be fielded by the USA and Sweden.
Lessons having been learned from the auto- GCAS effort, an auto-ACAS project was formulated to help prevent mid-air collisions. The device is only meant to be a "short-range system of last resort." Investigators are wrapping up development of architecture algorithms for the escape manoeuvres, and defining the required datalinks and sensors. An auto-ACAS will be built for flight tests on both manned aircraft and UAVs by 2003.
The auto-ACAS will provide aircraft with a safety sphere in which no penetration is allowed. It will determine whether one or both aircraft will accomplish an escape manoeuvre. The datalinks are required for in-network transponder-based co-operative operations, while sensors, such as radars, are necessary for out-of-network events involving non-equipped aircraft such as commercial transports.
Eberhardt says it is possible to design a sensor for the UAV that will detect out-of-network aircraft and use this information to initiate an escape manoeuvre. She also believes auto-ACAS technologies might eventually be used to reduce runway incursions.
The US Navy, Boeing and Engineering 2000, meanwhile, are taking a non-co-operative tack to solving the problem. The USN is funding an experiment to demonstrate a small, lightweight see-and-avoid system, combining passive and active optical and/or infrared (IR) sensors. Significant work remains to be done on the sensors, but a field demonstration of a prototype system is set for this year and early the next.
Engineering 2000, based in Woodinville, Washington, is developing a flight-ready, prototype optical system. It would use dual fish-eye optical imaging systems to identify potential collision threats. The system would point a laser radar (LADAR) at the threat to acquire range and closing rate. Yet to be determined is whether an electro-optical (EO) camera or an IR sensor will be used. TV cameras are inexpensive and offer high resolution, but clutter is an issue, especially in a look-down scenario. For IR, clutter suppression is more straightforward, but cost and weight are likely to be considerably higher.
The USN elected to concentrate on developing a small system that will eventually be flight tested aboard an Israeli-built AeroLite UAV, made by Aeronautics UAV Systems, and used by the USN as a testbed for small UAV payload and systems integration. The focus is mainly on the passive detection of potential threats in just the forward hemisphere, using range and range-rate measurements produced by a staring LADAR restricted to a relatively narrow search cone in the forward direction.
Dale Tietz of Austin, Texas-based New Vistas International is putting together an industry-government project to develop a non-co-operative, high-performance, all-weather, combined multi-mode system employing active (radar) and passive (EO/IR) sensors that could be fielded within two years.
The small forward-hemisphere-viewing gimballed system would leverage radar technology developed in the so-called Star Wars programme, and later adapted for helicopter obstacle avoidance systems made by Canada's Amphitech International. Passive sensors testing will be supported by US Government research organisations using fixed-wing aircraft and UAVs. Tietz also believes the sensor system could help alleviate runway incursions.
Navigating safely in civil airspace is one key technical issue that must be resolved. But the ability of fixed-wing and vertical take-off and landing unmanned air vehicles to make fully autonomous Category III instrument landings at civil airfields is also required if operations in civil airspace are to become routine.
Sierra Nevada has an extensive background in ground-based and aircraft carrier-installed landing systems for both manned and unmanned aircraft. A leader in providing automatic landing systems for UAVs, the company's UPN-51 Common Automatic Recovery System (CARS) is being integrated with various UAVs, including the AAI/IAI Pioneer, the General AtomicsRQ-1A Predator and the Northrop Grumman Fire Scout vertical take-off and landing tactical UAV. The tactical automatic landing system (TALS) was developed by the company for the US Army's AAI Shadow 200 tactical UAV.
The auto-recovery devices consist of a millimetre-wave radar sensor on the ground and a transponder on the target aircraft. The firm has developed control and tracking algorithms that allow the CARS and TALS to recover the UAV autonomously.
Meanwhile, the US military is funding development of the global positioning system (GPS)-based Joint Precision Approach and Landing System (JPALS) by Raytheon, which will allow for equivalent Category III instrument automatic landings by manned aircraft.
Seeing an opportunity, Sierra Nevada has integrated position data derived from differential GPS into the CARS, offering precision blind landings for manned helicopters and fixed-wing and vertical take-off and landing UAVs. It successfully demonstrated the expeditionary common automatic recovery system (ECARS) to the US Marine Corps in 1999, using rotorcraft.
"The ability to integrate ECARS avionics into fleet aircraft in a few hours, and to install ground systems in less than an hour demonstrated that ECARS is a viable solution to the USMC need for an expeditionary precision landing aid," says Sierra Nevada.
The company has gone one step further. It is transitioning the ECARS design into a dual-channel landing system for both manned and unmanned civil operations. Sierra Nevada believes an FAA-certifiable product can be created within two years. Candidate UAVs include the Global Hawk and Predator B. Company officials believe the all-weather autoland system will also enhance flight operations at oil rigs, heliports and austere airfields.
Looking ahead, several air traffic management initiatives now in development, such as automatic dependent surveillance-broadcast (ADS-B), could also help keep UAVs and manned aircraft apart. The FAA's Capstone project, now under way in Alaska, has taken ADS-B, GPS, datalinks, an onboard terrain database, graphical weather uplink services and traffic information broadcasts to create an integrated traffic/weather/terrain picture on a cockpit multifunction display, potentially offering huge pay-offs in navigation situational awareness. Some or all of these advanced technologies could be applied to solving the UAV see-and-avoid dilemma which stands in the way of the industry's growth.
"Without the ability to ensure that an equivalent level of safety [with manned aircraft] exists, commercial telecommunications and remote-sensing UAV operators will not be able to advance as fast as the consumer wants, thus delaying progress and affecting new business development on a global scale, concludes New Vistas' Tietz.
Source: Flight International
