UAV Autonomy - Flying sense

Washington DC
Source:
This story is sourced from Flight International
Subscribe today »

Jaws visibly dropped at an unmanned aircraft system (UAS) gathering in Washington in late April when the US Federal Aviation Administration's point man on the topic divulged his estimates for when the industry might have a certifiable detect, sense and avoid package for unmanned aircraft.

"I believe that the capabilities we're talking about will be available for 'file and fly' sometime between 2020 and 2025," said Doug Davis, director of the FAA's unmanned aircraft programme office.

Davis's estimate is the topic of much heated debate in the civil UAS community, where the military has campaigned for unfettered routine access - or "file and fly" - access to the national airspace system (NAS) by 2012 and where manufacturers would like to begin making business plans - and revenue - even earlier.

"I don't believe we can wait decades to fully integrate these types of systems into the NAS," said General Atomics Aeronautical Systems programme manager Sam Richardson at the same event, the first UAS safety forum held by the US National Transportation Safety Board.

Like other UAS makers, General Atomics, provider of large unmanned systems such as the Predator to the US military and NASA, wants expanded and simpler access to the NAS for training, homeland security and science missions, to name a few. General Atomics and four other UAS manufacturers have teamed as the unmanned air vehicle national industry team (Unite) to help accelerate UAS integration efforts under way at the FAA. The group's goals include creating "an environment" that supports viable public-use and civil unmanned aircraft applications that "operate routinely in the NAS" by 2012.

aai 
 © AAI

AAI Corporation Aerosonde UAV

With no regulations in place, public-use and military UAS operators must regularly apply for and receive a certificate of authorisation from the FAA to fly in national airspace. The approvals, issued on a case-by-case basis, generally last for a year and include numerous stipulations, including altitude, visibility and monitoring requirements, to ensure an "equivalent level of safety" to that of manned aircraft. The FAA expects to issue up to 100 certificates a year for the next five years, up from 85 last year.

Although considerable hurdles must be overcome in the areas of command and control, human factors, security and regulations before UAVs earn an equivalent level of access to manned aircraft, none are said to be more difficult, time consuming and expensive than developing an automatic collision avoidance system, also known as sense and avoid.

HOLY GRAIL

"Sense and avoid is the Holy Grail," says Rich Fagan, director of commercial UAS programmes for AAI, a Textron subsidiary. Along with its medium-sized Shadow 200, hundreds of which are deployed with the US Army and Army National Guard units, AAI also builds the small Aerosonde research UAV and makes structural components for Honeywell's micro-air vehicle (MAV).

That sense and avoid must be solved comes down to the FAA's insistence that UAS operations have an equivalent level of safety to those of manned aircraft, where sense and avoid is a basic ingredient in practically all operations. Asked at the NTSB meeting what he considered an adequate sense and avoid requirement, Davis responded: "One that RTCA presents to us when it's done."

RTCA is this case is RTCA special committee 203 (SC203), a government/industry consensus-based group formed in 2004 and tasked with creating minimum aviation system performance standards (MASPS), broad-based technology-agnostic rules, for sense and avoid and command and control.

Once the MASPS are complete, RTCA or another group will be tasked with developing the more specific minimum operational performance standards (MOPS) for sense and avoid and command and control. By following MASPS and MOPS, an aircraft or sense and avoid builder can greatly reduce the risk when developing a product that will pass FAA certification muster.

Given the stakes, SC203 is under pressure to sprint rather than crawl. When formed, the committee had considered 2009 as its goalpost for completing its work, says former FAA official and president of JSWalker Group consultancy, John Walker, who shares the lead role in SC203 with Ken Geiselhart of UAS-maker AeroVironment.

"We had two specific ad hoc groups come in and look at [the date]," says Walker. "The answer was 'no'." Walker says the committee discussed the evolution of traffic and collision avoidance systems (TCAS) for commercial aviation, a simpler technology than sense and avoid by comparison, but one that cost more than $1 billion to develop and nearly 30 years to institute. "What we're looking at is far more complex," he says. "It's more than collision avoidance it's separation assurance."

The FAA estimates that sense and avoid technology will cost around $2 billion to develop, according to a new study published by the Government Accountability Office.

Instead of the 2009 target, the group in March revealed a series of deliverable dates extending from 2010 through 2019 for the various MASPS, with sense and avoid coming last.

Pushing the schedule out a decade partly came from the group's decision to use a rigorous systems engineering process to develop the MASPS. Known as RTCA DO 264, the process was originally developed for establishing datalink services with air traffic control and was then modified for the UAS task by the SC203 committee.

After three years of preliminary work, the group has now started on the first of four backbone technical assessments - operational services and environment, operational safety, operational performance, interoperability - that are set to become the foundation of the performance requirements that will form the MASPS process.

INVESTIGATING OPTIONS

SC203 co-chair Geiselhart says the committee has investigated four options for accelerating the schedule, one of which will slide the sense and avoid deliverable by two years, to 2017. Another alternative will yield earlier draft or interim standards for "very specific missions and/or airspace classes," Ted Wierzbanowski, president of Unite, told Flight International. "This will allow us to develop, test, and safely gather the critical real world data on the performance of potential detect, sense and avoid and command/control system solutions as we strive to satisfy the near-term needs of both our military and non-military government customers."

Wierzbanowski says there are also calls for an independent review team to determine if there are "more optimum and/or efficient processes" than DO 264 for developing specifications. One of Unite's explicit goals is to "influence RTCA to adopt current available UAS standards and improve timelines."

NTSB board members questioned whether more manpower would help. But Geiselhart cautioned that "it's not just manpower" that is driving the schedule, but the need for "appropriate" skills of engineers, NAS operators, pilots, controllers, radio frequency specialists, manufacturers and others. "We've identified what we believe are the critical skills that have to be put on the project," he says. "Now we have to get commitments from the organisations that are part of this."

The juggling act for SC203 will be to balance the wide spectrum of stakeholder needs while developing standards that will be accepted by the FAA, an outcome that is not guaranteed. "We have a divergent stakeholder community. There are those from some UAS manufacturer perspective that have never had to deal with airworthiness certification of an aircraft operating in the NAS," Geiselhart says. "They want access to a market, but don't know what it takes to actually certify and introduce a new type of operational paradigm." At the other end of the spectrum are those who "want to be very certain that UAS doesn't put them at risk", he says.

Despite waiting potentially 10 or more years to completely define sense and avoid, Geiselhart says "point solutions" being developed today by manufacturers will not necessarily be moot. Included is Northrop Grumman's testing of a technology that will monitor and actively steer clear of as many as 30 "intruder" aircraft using sensors like TCAS, automatic dependent surveillance - broadcast, electro-optical cameras and purpose-built radar.

"What we're trying to do is establish technical standards that would allow manufacturers to present point solutions to the FAA for evaluation for airworthiness certification," says Geiselhart. "The burden and onus today for demonstrating compliance to the FAA is on the manufacturer themselves."

Crafting that helping hand will take time, however, a realisation that more and more industry officials are coming to realise.

When people get exposed to the complexity and degree of rigour that we're going through, it opens a lot of eyes," says Geiselhart. "But there are a number of fairly naive stakeholders who don't fully appreciate the engineering rigour and the regulatory certification processes that include public rulemaking. They represent the side of the spectrum that somehow thinks you can do this in three years."