[Peetwo]

Fawcett on drones, SMS & closing safety loops. 

Quote from the foreword of ICAO Annex 19:

The Standards and Recommended Practices (SARPs) in this Annex are intended to assist States in managing aviation safety risks. Given the increasing complexity of the global air transportation system and its interrelated aviation activities required to assure the safe operation of aircraft, this Annex supports the continued evolution of a proactive strategy to improve safety performance. The foundation of this proactive safety strategy is based on the implementation of a State safety programme (SSP) that systematically addresses safety risks.

Now a quote from M&M's SSP Chapter 2 on safety risk management:

..However, a modern approach to aviation safety management necessitates a systems approach to managing safety risks, encompassing organisational structures, policies and procedures—the SMS approach.


Safety risk management of the Australian aviation industry is a shared responsibility between industry and government aviation agencies. It is important that the aviation industry and the aviation safety agencies work collaboratively to produce the best safety outcomes.

The SSP recognises the need for a transition to a systems-based approach to safety oversight along with risk-based surveillance. This shift places more responsibility on regulated organisations and changes how regulators undertake oversight and monitoring roles.

The identification and management of aviation safety risk is undertaken through a multi-layered process which permits the aggregation of system and risk information into higher order categories, culminating in an assessment of the level of risk across the aviation industry.

---

And from M&M's SSP Appendix E:

...Aviation safety risk management in CASA


Consistent with the increased international emphasis on a state safety risk management programme, and as highlighted in ICAO Annex 19 (Safety Management) and ICAO Doc 9859 (Safety Management Manual), CASA adheres to the AS/NZS ISO 31000:2009 Risk Management principles and guidelines to effectively identify, evaluate, control (where CASA has risk ownership) and monitor aviation safety risks.

As outlined in Chapter 2—State Safety Risk Management, management of Australian aviation safety risk is undertaken through a multi-layered process that has the capacity to identify and manage risks at various levels of the aviation industry...

There is at least one person in Australian politics that fully understands the principles and advantages of an effective aviation safety management system in proactively addressing identified safety risk issues, that person is Senator David Fawcett...

 Last week we were witness to Senator Fawcett's deeper understanding of an SMS and the important role the Government's aviation safety agencies are proactively supposed to play to ensure the integrity and effectiveness of the Australian version of the ICAO Annex 19 SSP.

Unfortunately it would appear that these agencies still lack the insight to understand the holistic, all inconclusive approach required to turn the SSP into a document of more than just words...

In the following Hansard excerpts/Youtube vid,  Senator Fawcett proves that these agencies a seemingly content to stand back and watch the drone related accident occur, instead of collectively working together to proactively work on solutions, as much as possible, to mitigate the risk:       

Senator FAWCETT:  My last question goes to the issue of risk. You've talked about risk assessments. As you do your risk assessments are you relying on CASA's evaluation, which essentially supports their two-kilogram threshold? Or are you aware of more recent studies, for example from the UK and their Military Aviation Authority, that indicate, certainly for GA aircraft and for rotor craft, that much smaller drones present still catastrophic risk?

Dr Weaver : I'm aware of and have read that study from the UK. Our approach is to take all data sources on board. In some respects once somebody has entered controlled airspace we look at how we manage the risk, no matter what the size of the aircraft or the drone is. So, our risk assessment is focusing on what we can do. Obviously it will also focus on what we can encourage other parts of government and industry to do as well. But we're targeting it down to our role, our controls and our mitigations that we can put in place.

Senator FAWCETT:  Lastly, if we were to go to a model where we have a whole-of-government approach where the importation and sale is limited to the kinds of systems that DJI are putting out with that off-the-shelf, very limited operating envelope, and then we have a qualification regime where people can expand that a little bit to what our current requirements are—no closer than three kilometres, no more than 400 feet et cetera—and then another one to a commercial standard, from Airservices' perspective, who do you see should actually be conducting that training and licensing? Is that something that you see should be done by CASA, as it is for air crew at the moment? Could that be done by a commercial provider? Who do you think should be doing that?

Dr Weaver : When I look globally, I see a range of alternative approaches that occur. You see commercial providers of drone training internationally, and then you see some where it's more regulatory based. I guess, as the air traffic service provider, I'm agnostic as to how that's implemented. I'm interested in the outcome of preventing access to controlled airspace.

CHAIR: I want Senator Fawcett to have as much time as he needs. But I want to come back to estimates of 23 May—a Tuesday—this year when we had sitting at the table Mr Chris Manning, a former Qantas chief pilot who is now commissioner at ATSB. The words ring in my head. Although, I can't say them verbatim. Commissioner Manning said to us, 'Anything in airspace is a concern.' I get all that. So I want to ask the experts. We haven't had an incident—which is great. That's lucky. We don't want one. But, Mr Nagy, when you hit the pelican is there any way that something like this could do more damage to a fixed-wing aircraft?

I'm not going to won't worry about helicopters because that's Senator Fawcett's area of expertise. Would or could it do more damage than a birdstrike if it hit the engine or something like that?

Mr Hood : If I may, I would like to table this, an ATSB report where a wedge-tailed eagle, weighing about four kilograms, actually went through the cockpit of an Glasair aircraft near Bathurst in 2015. The pilot was temporarily blinded and broadcast a mayday. He had dead bird throughout the cockpit and in his lap, so the bird actually did go through the windscreen—a four-kilogram bird went through the windscreen. If you equate that to a four-kilogram RPA, obviously that is possible in light aircraft. Am I able to table that, please, Chair?

CHAIR: I can appreciate that, because I know a bloke who was minding his own business north of Pardoo one night when a bush turkey came through the bloody windscreen, mate—I've got a tell you, apart from crapping myself, lucky I wasn't at 30,000 feet.

Mr Hood : Apologies to Senator O'Sullivan, I will get a copy of that report sent to you electronically.

Senator O'SULLIVAN:  Thank you.

CHAIR: We've heard evidence about lithium batteries and all that. I want you to tell me. Do we just say, 'Oh, if it's a bird or one of those, it doesn't make any difference—it's the same'?

Mr Hood : I think this is where we don't understand enough. This is where we're very interested in the UK report and what the difference is between striking an RPA and striking a bird. We have about 2,000 bird and animal strikes per year in Australia. Most of them bounce off to little effect, but, of course, the one in the report in front of you did enter the cockpit, breaking the canopy.

Senator FAWCETT:  I think the UK report is very explicit on that. It highlights that the degree of give in the plastic of a plastic-covered drone has a very different impact to, perhaps, the homemade one where there are exposed metal components. So with the bird analogy, with lots of soft flesh tissue around, it will have an even more absorbent impact than the plastic. The UK report's actually pretty definitive on that point. A drone with exposed metal components, even at 400 grams, will have significant impact on a helicopter when it comes to dynamics components but even on a GA aircraft. That brings me to the issues around likelihood and consequence, which underpin risk. It brings me to issues of compliance and culture and, lastly, capacity. All of that will lead me to question you, as one of the safety experts, as to what work you are doing to contribute to a systems safety approach to this issue, as in a whole-of-government approach.

In terms of likelihood, Mr Holman, you very dutifully gave the answer that a DJI can only fly to 400 feet because that's the rule, and we had evidence here that you can't fly beyond line of sight. Can I tell you, the culture amongst the operators is different. Go and have a look at the DJI Forum online. These are people who fly DJIs for fun. A post earlier this year said, 'The FAA's 400-foot rule is not a rule; it's only a guideline.' Some bush lawyer has gone through and looked at all kinds of regulations going back into dim, dark history and statements in congress et cetera to justify why people can ignore the 400-foot rule. That's the culture. What that says is that, despite CASA's app and despite the piece of paper they put in the packet, the live culture that is dictating how the people among the 49,000 that the chair is concerned about operate these devices is that the likelihood of them being in places where they should not be is high and probably growing. It's also growing because things like the DJI off the shelf are more and more capable with every evolution. Therefore, there are more of those 49,000 who are flying more high-capacity aircraft. Putting that likelihood together with the consequence that we've seen from the British report, which you've very usefully highlighted again for us, what's your assessment of risk now compared to perhaps a year ago?

Mr Godly : The British report definitely provided a lot of information to us. Six months ago, when we were talking at the estimates committee, there wasn't really any of that real research, just mathematical models, basically based on birdstrikes and not much else. So that report really has informed us and what it has informed us is fairly consistent with what the models were suggesting would happen. In particular, for airliner-type aircraft, the risk of a collision—in particular, the windscreen—is probably limited to that cruise speed, because at the approach and descent speeds the bird strike certified windscreens don't seem to be compromised. And when you take into account the likelihood of that—we are seeing that three per cent of our encounters are above 10,000 feet, but the majority are obviously below that. So when you put likelihood and consequence together, for at least the high-capacity aircraft it's probably a fairly low risk.

For general aviation it's probably a different story. They generally don't have bird strike certified windscreens. And as the British report showed, there's a much greater chance of penetration of a windscreen, and also things like damaging the wings and rotors of helicopters. So in my opinion the biggest risk is for general aviation aircraft.

Senator FAWCETT:  Well, perhaps I can put to you, because your submission highlights this, that the percentage of incidents for rotor craft as a percentage of hours flown shows that they are actually probably in the zone of having the most likelihood of an interaction. And perhaps I can also put to you that unlike a strike on the wing of a GA aircraft, which may put a dent in it or may even rupture a fuel tank, in worst case, a strike on a tail rotor of a helicopter probably is going to mean the loss of that helicopter. Whilst the military used to train, somewhat hopefully, to teach people to land a helicopter without a tail rotor functioning, and the centre-of-gravity impacts and rotational impacts et cetera, I'd argue that the majority of your pilots would probably struggle to land a helicopter without a tail rotor, which means you've got whole loss and life loss. So, the consequence is incredibly high. The consequence for your RPT aircraft, whilst it may not crash—we've already seen at Gatwick, incidentally, on 2 July, the same day we had the Fokker incident in Brisbane, aircraft having to divert, or hold, which means cost disruption—other than the windscreen, you've got two or maybe four very large intakes to engines. The cost of one of those going through an engine would be very large. So whilst we're not talking about life necessarily, there is a cost and a consequence across the whole range.

Coming to recognising, as you said in your submission, your priorities and your capacity to investigate things, CASA likewise has capacity constraints. Does it not suggest, if we have high risk and capacity constraint within a regulator like CASA and organisations like yours, that we need a whole-of-government safety system approach to this which includes a prohibition on the import of devices that are capable of these kinds of operations unless they are, as one previous witness told us, off the shelf, geofenced to a very small bubble—100 or 200 feet—around the operator, with increases to that operational capability allowed only when they demonstrate that they have complied with a level of training? And I would argue that the first level of training is an independently invigilated exam that shows that they understand the safety implications of the current envelope—three miles or three kilometres around airports, 400 feet et cetera—and that the geofencing is expanded to a hard limit of 400 feet and the other area and then a final level, which is commercial, all the normal commercial considerations that currently exist, would remove those limitations from the device. That kind of whole-of-government, importation, customs-type restriction as well as CASA, as the regulator, surely, given your capacity constraints, is the sort of model we should be looking at. Have you done any work as the premier safety body to look at a systems based safety approach involving the whole of government as opposed to limiting your thinking to the current aviation environment—which is yourselves, Airservices and CASA?

Mr Hood : As I mentioned before, we're informers of policy and we're working very hard on this issue to ensure that we inform those we are required to inform. Certainly under the CASA ATSB MOU we're feeding the stats and our analysis of those stats to the Civil Aviation Safety Authority and the department. We have also added RPAS to our safety watch. We have nine areas of focus—safety watch, that's called—and RPAS is certainly one of those. We certainly are gaining our own experience in relation to the operation of RPAS—for example, the grounding on our very first sortie, if you like, to use the drone for investigation: on the train, heard the helicopter, let's put it on the ground.

So it's kind of a learning experience for us, being an operator ourselves in the RPAS environment. Our view is that we are working very hard to ensure that we also inform the travelling public of the emerging risk. I did take out something from the previous discussion with the Airports Association that maybe we could do a better job of highlighting those maps that we put in our submissions in terms of local councils and airports, to say, 'These are the near encounters in your area.' Potentially that would inform them in relation to the signs. But I know what you're saying—the policy objective: we're informing the policymakers.

Senator FAWCETT:  But what I'm asking is: are you looking for broader policy recommendations than perhaps you have traditionally taken? You know as well as the rest of us do that ministers tend to respond to recommendations of departments. If they go beyond recommendations of departments then, as Sir Humphrey says, 'That would be a very brave, Minister'! So, if departments come only with recommendations that are constrained by their current thinking—and from evidence that we've heard in this inquiry to date it seems that nearly every agency is constrained by the current construct of their role—CASA tells us, 'Well, we don't control importation', and you tell us you don't control regulation and Airservices says the same, who is doing that whole-of-government systems-type thinking to bring a recommendation to the minister saying, 'There is a way we can tackle this, but it will involve cross-departmental action and policy that brings in a number of arms of government'? I'm not hearing from any of you that anyone is doing that level. And if you, as the nation's premier safety experts, can't approach this from a broader systems perspective, asking what are all of the levers of government that we can bring to bear to this, then on what basis can we expect the minister to take action?

Mr Hood : I might make a couple of points. Australia does have an aviation policy group. The ATSB is not a member of the policy group, because it may well be that we have to investigate policy decisions. The department, Defence, CASA and Airservices are all members of the aviation policy group in Australia.

The other point I might add is the ATSB is very much the canary in the mine—and let me tell you, we will sing. But we're not going to sing prematurely, and we're not going to sing without the evidence to sing appropriately. You may be aware, from the other committee, of our report on the ATR aircraft. We have formed a strong view in relation to that aircraft. We published two interim reports and we've got a third one coming. So, we're not afraid to exercise that authority and have our say when we think it's appropriate. In this particular case I suppose it still comes down to the fact that we don't think we have enough evidence yet in relation to the consequence.

Senator FAWCETT:  Okay. Thanks, Chair.

---

---

Senator FAWCETT: Sorry. You've completely derailed my train of thought, Mr Crawford.

CHAIR: I've been trying to do that for the last six years you've been here and I've never succeeded!

Senator FAWCETT: You have. You'll have to take some lessons. CASA has previously expressed some concern about the maturity of technology such as geofencing. I've actually just been going through your submission again, and I thought you'd mentioned something in here that some of those technologies could potentially introduce risks. But we heard a comment before about DJI as an OEM—and I haven't been able to clarify whether this is their intent or whether this is what they are doing now—and that their product off the shelf is limited to a 100-foot bubble around the operator. If that concept is viable and mature enough that an OEM is doing that, why would we not look at a whole-of-government approach where we limited imports and sales to only OEMs that were prepared to take that approach so that we completely avoid the example of the Christmas present with an ill-informed operator who happens to fly it under a helicopter route or near an airport? And if it is true that DJI have put this in place, then it says that it's mature enough that it's commercially viable, therefore it's probably reliable enough that we can actually start limiting the 90 per cent of the unintentional incidents due to lack of knowledge. Why would we not take that as a key approach to the government?

Mr Carmody : Firstly, I'm not sure that it's true. One of my colleagues might know, as we stay as closely as we can to these sorts of developments. If geofencing—and I'll call it 'geofencing' in that context—or limiting a bubble around a drone is technically feasible, does mature and does become that way, that is certainly one of the methods that you could use to control drones and manage some elements of the risk. I'm not certain that it's as mature as advertised as yet. They're a very big marketer of drones—the biggest in the world. They're obviously trying to stay the biggest in the world, or get bigger. I'm not sure how it actually interacts with other technologies as well. So, I think that the jury is out. It sounds logical and sensible, and we will certainly consider it. But in terms of whether it is there yet, I'm not sure.

Senator FAWCETT: Could you undertake to have one of your people contact them, and OEM, and ascertain and come back to the committee with a view on that.

Mr Carmody : Certainly.

Senator FAWCETT: Because, if it's technically feasible and we limit the market to only manufacturers who meet that technical bar, as sure as apples come from trees, other manufacturers will reach that bar if they want to sell into the market. That then gives us a starting point where, over five years, with degrading batteries and all the rest of it, the old fleet will disappear and we'll be in a much better space for those 49,000 recreational users to encourage, through capability, their requirement to increase their level of knowledge before their machine becomes more capable.

Mr Carmody : We certainly will, Senator. We'll certainly take it on notice. We'll ask the question. I was hopeful we might even have the answer, but I assume we don't. We'll ask the question and come back to the committee and let you know what we find. I think, conceptually, in isolation, it sounds like a very positive outcome. I would just like to see how it fits.

---

Senator FAWCETT: Don't get me wrong. I applaud the range of efforts in issues like the cinema and things like that. It's fantastic in terms of informing people. But I think you were here before, when I talked about the DJI blog site, or the forum. Despite all of the training and licensing that we give to professional pilots, you get the occasional person who thinks it will be fun to beat someone up or push the boundaries et cetera. If you have a large population who have never been through that rigour, nor understand the risks that emanate from their conduct, then that kind of culture that's reflected in the blogs says that trying to regulate and educate through what is essentially a voluntary system is going to be almost unmanageable. That is why I think we need to have a system where we use the technology to limit the exposure to other aviation users, and only allow it to expand as you do at the moment. The commercial RPAS licence regime is very thorough, and it's great for people who want to operate commercially. What's missing is the middle piece between someone who can essentially buy a toy that they can fly to 100 feet around them and someone who wants to operate something with more capability.

This is my final question for you. Let's say we did go down that three-tiered system, where, without constraint, you can buy a toy, but you've got a technical constraint that you can only fly it in a little bubble, and, at the other end, we already have your commercial system, but in the middle there's the technology piece—and we'll explore that—but then there's the education and licensing approach. I've raised before the example of the maritime radio operators' licence, where, dangerous as that device is, every user has to actually pass an independently invigilated exam. Would CASA be the right body to take your current rule set—around three nautical miles, 400 feet et cetera—and create an online course or syllabus that perhaps could be run by the private sector and then an exam that people would sit so that they could then, with a number saying, 'I have demonstrated a degree of knowledge and competence by passing this exam,' go back to the OEM to get that technical unlock to operate in that middle recreational space? Would CASA be the person to do that?

Mr Carmody : You could do that, in reality. It is achievable. You've gone through quite a number of steps. There's a fair bit of complexity in that, I might say. But yes. We are the regulator. It is fundamentally our responsibility. The question would be—and that's one of the things that we're looking for in our survey—how much education would be enough? How much do you actually need? As you indicated, quite rightly, in the commercial sector there is a significant amount of rigour. How much, verging from nothing to that, do we actually need? And how would you manage such an arrangement? And how would you link it if the government decided to register every drone, for example? How would you make all of these things fit together? As I said, I think there is a fair bit of complexity in this space. But, on its face, it's achievable. But it would cost.

MTF...P2