Category: Flying

AI2

There’s not just one form of Artificial Intelligence (AI). This group term hides a great panoply of different configurations, shapes, and forms of applications.

One of the most impactful applications is that of machine learning or expert systems. It’s where we go beyond a conventional computer’s ability to store and manipulate information against set rules. It’s where the machine has the capability to learn new ways of interpreting information and thus becomes different every day of operation from the day it was switched on. That’s a bit vague but it captures the essence of moving from deterministic to non-deterministic systems.

In all this we do presuppose that such complex systems are in the hands of able and highly illiterate users who understand what they are doing in training that learning machine. There’s debate about how bias in algorithms can produce unintended consequences. In addition, a reliable and trustworthy machine can be trained in a way that embeds errors and biases too[1].

Just as a child picks up the bad habits of a parent, so “intelligent” machines can learn from pilots, controllers and engineers who may have less than optimal ways of undertaking tasks. This Human-AI interplay is likely to become a major area of study. As the topic of Human Factors is itself a large body of material.

Already with the debate on social media it is all too obvious that the aviation community has a wide range of views on the use of AI. All the way from utter rejection, or scepticism deeming such systems as “unsafe” to advocates who profess only the benefits and merits of such systems.

Clearly, both extreme ends of the spectrum of professional views don’t help much. I don’t think that the promoters of AI want to see blind overreliance on it. Equally, surly even ardent sceptics can see virtue in making the best use of the accumulated knowledge that is available.

I can foresee a system of systems approach. With my parent and child analogy, from time to time a child will ask a question that is blunt and to the point. A question that demands a straightforward answer. This can be uncomfortable but hits out at biases and bad habits.

In aircraft systems there are boundaries that must be respected. The physics of flight dictate that going beyond those boundaries is generally not good for life and limb. So, a system programmed to question an expert system, one AI questioning another AI, or even question its trainer, is not beyond the realms of possibility. It might even be a good idea.

[1] https://www.nature.com/articles/s41746-022-00737-z

Poor law making

If you thought the Truss era was an aberration, and that the UK’s Conservative Party had learned a lesson, then please think again. Wheels set in motion by the ideologue Jacob Rees-Mogg MP are still spinning.

The Retained European Union Law (Revocation and Reform) Bill is trundling its way through the UK Parliament. The Government Bill will next be prepared for its 3rd reading in the House of Commons[1]. The Conservative Government has brought forward this Bill to revoke, reform or revise all the remaining law in the UK that was formerly derived from the UK’s membership of the EU. This turns on its head the normal approach to changing UK legislation. Revocation is automatic unless there’s an intervention by a Minister.

UK civil aviation depends on several thousand pages of legislation derived from EU law[2]. Much of this law was created with considerable contributions from the UK. There’s hardly any if any advocates for automatic revocation of current aviation legislation. Even the thought of this action sends a shiver down the spin of aviation professionals. Generations of them have worked to harmonise rules and regulations to ensure that this most international of industries works efficiently.

Unless amended, the Government’s EU Retained Law (Revocation and Reform) Bill[3] could turn out to be an absolute disaster. Even those who have an irrational wish to eliminate any and every past, present, or future link to Europe must come up with a practical alternative and do this in an incredibly short time. Without a consistent, stable, and effective framework civil aviation in the UK will grind to a halt. Again, even those who have an unsound need to change for change’s sake will be hitting a vital industry hard, as it is only just getting back on its feet after the COVID pandemic and now setting out to meet tough environmental standards.

It’s going to be interesting to see what happens when this poor Bill reaches the House of Lords. Once again, the country will be relying on the upper house to add some common sense to this draft law.  

POST 1: The 3rd reading debate makes it clear that the Government is unsure which laws are covered by the Bill. If the Ministers responsible for this legislation do not themselves know its extent, how can anyone expect civil servants working on this legislation to know the full extent of change? A most strange state of affairs Retained EU Law (Revocation and Reform) Bill (Third si – Hansard – UK Parliament

POST 2: Retained EU law lays down rules for the airworthiness and environmental certification of aircraft and related products, parts and appliances, as well as for the certification of design and production organisations in the UK Commission Regulation (EU) No 748/2012 of 3 August 2012 laying down implementing rules for the airworthiness and environmental certification of aircraft and related products, parts and appliances, as well as for the certification of design and production organisations (recast) (Text with EEA relevance) (legislation.gov.uk)

POST 3: Retained EU Law Bill is being debated in the House of Lords on Monday, 6 February.

[1] https://bills.parliament.uk/bills/3340

[2] https://www.eiag.org.uk/paper/future-retained-eu-law/

[3] https://www.gov.uk/government/news/the-retained-eu-law-revocation-and-reform-bill-2022

SPO 3

Now, there’s an activity with two humans in the loop. Given the physics involved the goalkeeper should be beaten every time. Well, I’m saying that assuming a high level of expected performance on the part of the footballer taking the penalty. I guess that’s why we are often critical when they miss. In the last few weeks there have been more than a few examples to watch.

What we know is that football penalties are much more than mechanical actions and reactions. However, there’s a degree of mythology about the inevitability of human factors taking control of the outcome: goal or no goal. I’d like to think that there’s an ever-shifting blend of what physics does to the ball and what the human does. Is it always possible to predict the slipperiness of a spinning ball traveling at speed that is then touched by the fingertip of a goalkeeper?

What if the footballer taking the penalty, was an “intelligent” machine. That is a machine with a sensor array and computational capability that far exceeded normal human performance. Such advance automation could calculate the most probable reaction of a goalkeeper based on history and the immediate movements they make right up to the last millisecond before the ball is struck.

Assuming the machine was limited in term of the force it can apply to the ball, it could still adjust its actions as soon as any new information was available. I’m not saying the outcome will always be better for the machine football striker. However, it could reduce the scope for error and randomness to dictate what finally happens.

So, with that argument, in aviation, I’m saying it’s not right to say that Single Pilot Operation will always be worse than two crew operations. Don’t get me wrong, those people aggressively advancing the idea that the intelligent machine will always be better than a human are missing something too.

One thing that highly capable automation could have to bring to the party is not only early detection and diagnosis of problems but a massive library of stored experience. How we embed and constantly update that flight experience is an almighty challenge.

Afterall, the dread in aviation is knowledge with hindsight. It takes the form: “You should have known. Why did you let this incident happen?”

I’m now tempted to think of a Star Trek analogy. Every second an aircraft of a type is flying, experience of its operation is being accumulated. If there are hundreds of a type flying at any moment across the globe, that’s a lot of data to collect and absorb and think about before acting. 

The fictional and scarry Borg are cybernetic creatures linked by a hive mind and they know a thing or two about assimilation. Granted that’s farfetched as analogies go but my point is that I believe we are generations away from that kind of capability. Not only that, just as humans fail so any such “intelligence” designed by humans will fail to.

SPO 2

An instant reaction to Single Pilot Operations (SPO) is like the instant reaction to completely autonomous flight. “I’m not getting on an aircraft without a pilot!” Then to justify that reaction fatal accidents of the past are cited. Typically, this is to remind everyone of the tragic Germanwings accident[1]. It was 24 March 2015, that an Airbus A320 was crashed deliberately killing all onboard.  

However, it’s wise to remember that the likelihood of incapacitation[2] is much greater than that of the malicious behaviour of the pilot in command. Cases of malicious behaviour leading to a catastrophic outcome are truly shocking but extremely rare.

One fatal accident, that is still disputed is EgyptAir Flight 990[3] that killed 217 people in 1999. The possibility of inflight pilot suicide is unnerving, since on the face of it there is little any of the aircraft’s cabin crew or passengers can do to stop it.

This could be a future opportunity to use automation to prevent these scenarios occurring. Afterall the aircraft knows where it is and that a sustained high-speed dive towards the ground is not normally intended. A safety system exists to do this[4], but its outputs are not connected to the aircraft’s flight controls.

Humans being adaptable, extremely creative and capable of highly irrational actions, it’s unlikely that malicious behaviour resulting in aviation accidents will ever be reduced to zero. This is said regardless of the procedures or technology involved. The fate of flight MH 370 remains a mystery.

Thus, the prominent safety issue in respect of SPO is pilot incapacitation. Where the pilot in command is no longer able to perform as expected. That is, if the aircraft flown is not capable of safely landing itself. The objective always being safe continued flight and landing.

I’ve had the “1% Rule” rule explained to me by a notable aviation doctor, but I must admit I didn’t fully take it in. So far, the rule has stood the test of time. When the pilot in command of a Czech Airlines aircraft collapsed and died on route from Warsaw to Prague in 2012, the co-pilot took over and everyone got home safely.

Any automated co-pilot must be at least as capable as a human co-pilot in all aspects of operation of an aircraft. The key word here being “all”. It’s not enough to have the functions necessary to undertake safe continued flight and landing. Task such as communicating with the cabin crew and passengers must also be considered. Including preparation for an emergency landing.

[1] https://www.bbc.co.uk/news/world-europe-32072218

[2] http://www.avmed.in/2012/02/pilot-incapacitation-debate-on-assessment-1-rule-etc/

[3] https://www.theguardian.com/world/2002/mar/16/duncancampbell

[4] https://skybrary.aero/articles/terrain-avoidance-and-warning-system-taws

Single Pilot Operations

Single Pilot Operations is not new. What’s new is considering this way of working for everyday public transport operations of large aircraft

Research is of fundamental importance. It seems obvious to say so given the benefits it has given us. When proposals come forward to exploit new technologies there needs to be that moment when everyone steps back and takes a long hard look at the implications of its use.

In basic technical research it’s not the most important consideration is to focus on the drivers for change. They can be multifarious: economic, environmental, social, safety, security, political, and maybe just a matter of preference. Policy directions are taken by the industry and governments not constrained by what is happening now as much as what might happen tomorrow.

Research has delivered incredible safety improvements in aviation. This is not only in the basic design and construction of aircraft but all aspects of their operation. So, to see that the European Union Aviation Safety Agency (EASA) sponsoring research to study the implications of aircraft Single Pilot Operations[1] is a wholly good measure.

My history goes back to the early days of fly-by-wire aircraft systems. This is where the mechanical and physical connection between an aircraft pilot’s actions and the control surfaces that determine flight are replaced by digital computers. Back in the 1980s, a great deal of research and experimental flying proved the technology to make fly-by-wire work. It first found favour with the military. One reason being that an aircraft’s capability could be extended well beyond what was formerly reached. This change was introduced with caution, analysis, testing and much detailed risk assessment.

At the time, there was a significant body of professional pessimists who predicted a diminishment of aviation safety. Today, four decades on, studies show that even as air traffic has increased so civil aviation safety has improved. A momentous achievement. An achievement that has, in part, been because of the well-regulated adoption of advanced technologies. 

It is important to look at potential changes with an open mind. It’s easy to come to an instant opinion and dismiss proposals before a detailed study has been conducted. The detailed technical research can then be part of the challenge and response that is necessary to before approval of any major change. First difficult questions need to be tabled and thoroughly investigated.

[1] https://www.easa.europa.eu/en/research-projects/emco-sipo-extended-minimum-crew-operations-single-pilot-operations-safety-risk

Air Safety List

A long time ago in a far away place. Well, that’s how it seems, and it was more than 17 years ago.

A flight ban was placed on Turkish airline Onur Air back in 2005. At that time, I was in my first full year in Cologne, Germany building up the European Aviation Safety Agency (EASA). We were well on the road managing the handover of responsibilities from activities of the Joint Aviation Authorities (JAA) to EASA. However, the European legislation that empowered EASA was in a first and most basic version. This was planned to be so because taking on aircraft certification work was a big enough task to start the new Agency.

The JAA had coordinated an aircraft ramp inspection programme and maintained a centralised database for its members. This was where a member state would inspect an aircraft arriving from a third country to ensure that international rules were fully met. The SAFA programme was launched by the European Civil Aviation Conference (ECAC) in 1996. SAFA standing for Safety Assessment of Foreign Aircraft.

Onur Air failed such inspections, and the Dutch government imposed a flight ban[1]. Similar bans were imposed by Germany, Switzerland, and France. However, if my reflections are correct the airline moved operations to Beligum where there was no ban. As you might imagine this caused concern amongst EU Member States. Where everyone had agreed to cooperate on aviation safety matters there seemed to be a degree of incoherence.

Long before the first EASA Basic Regulation, which by the way, didn’t address this subject, there was Regulation 3922/91[2]. I remember a hastily convened committee composed of representatives of the Member States and chaired by the European Commission (EC). The “3922[3]” committee hadn’t sat for years but then it sprung into action in response to the lack of a consistent approach to airline safety bans across Europe. I was there representing EASA.

So, the EU Air Safety List was born and the associated legislation[4] to support it. Even though the UK has left the EU, and left EASA this safety list remains the basis of the UK’s own Air Safety List[5]. Adding and removing air carriers and States that fail to meet internationally agreed safety standards is work that no one State should do alone.

[For safety’s sake, this should not be one of the parts of adopted EU legislation the UK Parliament wants to sweep away with its planned new Brexit law].

POST: Current list The EU Air Safety List (europa.eu)

[1] https://www.expatica.com/nl/general/dutch-lift-ban-on-onur-air-38258/

[2] Council Regulation (EEC) No 3922/91 of 16 December 1991 on the harmonization of technical requirements and administrative procedures in the field of civil aviation.

[3] https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ%3AL%3A1991%3A373%3A0004%3A0008%3AEN%3APDF

[4] https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32005R2111&rid=6

[5] https://www.caa.co.uk/commercial-industry/airlines/licensing/requirements-and-guidance/third-country-operator-certificates/

Bird Strike 2

What makes a bird strike different is that it’s an unpredictable collision. If we talk about aircraft collisions with terrain the outcome is predicable bad. All that kinetic energy must go somewhere. So, a high-speed vehicle hitting something that is immovable is not going to end well.

Now, it must be said that some hunting birds can dive at incredible speeds. More typically, a large bird in flight between feeding sites isn’t going to be traveling fast. In fact, it may as well be viewed as a static object relative to an aircraft in flight. A bird in-flight is unlikely to be able to take avoiding action. For a pilot the action of “see and avoid” may work in respect of other aircraft but a bird ahead is no more than a pinprick in the sky.

These factors make aircraft bird strikes inevitable. That said, the range of outcome because of impacts are rarely at the severe end of the scale. One reason for this is the effort made at design certification to ensure an aircraft is sufficiently robust. Damage can occur but if the aircraft design and test processes have been rigorous everyone should get home safely.

I remember paying particular attention to the zonal analysis done by several major manufacturers. In my experience the most difficult designs are for those of business jets and large helicopters. One of the design challenges in both cases is the limited physical real estate within the aircraft structure. Weight is another big consideration. This leads to cramming essential avionics and electrical systems and their interconnections into confined spaces.

Zonal analysis is about ensuring there’s segregation between different systems. Afterall what’s the point of having two Attitude & Heading Reference Systems (AHRSs)[1] and putting them next to each other in the nose cone of an aircraft. That’s not a good design strategy. One damaging impact must not take out two essential independent aircraft systems.

It’s just as important to ensure an aircraft’s wiring isn’t all bundled togther and taken through one connector. That may save money on electrical parts but it’s not going to work after being hit hard by a 5kg goose.

These issues will need particular care in the new electric vertical take-off and landing (eVTOL) aircraft that are on the drawing boards. Choosing a safe architecture, manufacturers must balance the use of creative design solutions, to produce a competitive product, with limited physical space.

A couple of key words in the certification requirements concern hazards that are anticipated. Bird Strike is hazardous and aircraft systems and equipment must “perform their intended function” should it occur. See EASA Special Condition for small-category VTOL aircraft, Subpart F[2].

POST: Good to see the bird strike criteria Joby’s airworthiness criteria: A blueprint for the nascent eVTOL industry – Vertical Mag

[1] https://helicoptermaintenancemagazine.com/article/layman%E2%80%99s-guide-attitude-heading-reference-systems-ahrs

[2] https://www.easa.europa.eu/sites/default/files/dfu/SC-VTOL-01.pdf

Practicable

I am in two minds. When I see the words: “in so far as practicable” I can think; great, a degree of flexibility. On another occasions when I see the self-same words I think; that’s too vague and indistinct. It can easily circumvent more strict language.

Practicable is a perfectly useful word. The idea that it’s practical to do something and likely to succeed can be a matter of reasoned judgement. However, there lies the crux of the problem. It’s the subjective of that judgement, as to what’s practical and will it succeed, that becomes the possible difficulty.

If there’s a clause in a group’s constitution or working arrangements that says: “in so far as practicable” then it can become open season for someone to avoid a commitment or go their own way. That can be to shoot a big hole in a set of agreed expectations.

A lot depends on where the burden of proof sits. In other words, I may assert that something is not practicable but is it then for someone else to prove me wrong? Or do I have to provide the necessary proof?

There are elements of degree here too. If the assumption is that a judgement can be a snap judgement that’s one thing. However, there may be an assumption that a judgment is based on a rigorous level of analysis and reasoning.

The term “in so far as practicable” is most useful when applied thoughtfully and with honest intent. That the person applying this caveat would work hard to undertake whatever obligation was written and only as a fall back, having been unable to meet an obligation, revert to the use of these words.

Context and circumstances weigh heavily on what is practicable. An easy task on a sunny day can be a nightmare in a thunderstorm. Some legal clauses go as far as “insofar as is reasonably practicable in the circumstances”.

I guess I’m coming around to the wish that the “ALARP” concept (short for “as low as reasonably practicable”) would be sparingly used. ALARP weighs risk mitigation, elimination or reduction against time, trouble, and money. That’s a balancing act where there’s no perfect answer.

ALARP is a basic concept in health and safety law, but it hasn’t caught on in aviation or at least safety of flight. It’s not that aviation is blind to the fact that flight safety can be a priority, but it will never receive infinite time, trouble, and money.

It’s more that with flight there’s always a choice. It’s a “go – no go” choice. If adequate risk mitigation, elimination or reduction is not available the reasonable choice is to stay on the ground.

Luggage

It’s a space we have control over. Not a house or a room but, most often, a volume of space no greater than what we take up in our human frame. It’s not organic. It’s far from that because its role is security, storage and logistical. That’s the humble suitcase, and a great array of bags and backpacks that help us get from A to B with enough possessions to make life comfortable.

The choice of a suitcase or bag is not a trivial matter. Lessons from experience range from bursting zips to leaking contents that turn favourite clothes into damp rags. The challenge of replacing a cabin bag or case takes research and careful weighing of multiple options.

If traveling by air, there are numerous constraints on size and weight. A completely free choice as far as colour is concerned but that’s about the only characteristic that’s an open book. That said, it’s astonishing how many black cases look like other black cases in the array of black cases.

More than a decade ago airlines started charging extra for hold luggage on top of their basic fares. Since then, flying with hand luggage only has become popular. This trend can be troubling. Watching passengers squeeze unreasonably sized bags into overhead bins is not an entertainment. The expectation that an aircraft overhead bin can take a massive bag is not a reasonable one.

My latest purchase has been made from recycled plastic bottles. Naturally, that conveys a fell good factor. It’s a great way to give new life to the huge numbers of discarded single use plastic bottles that somehow we’ve become dependent upon. In my childhood, I don’t remember any plastic bottles. Plenty of glass but no plastics.

For short journeys, the faff of checking-in a suitcase, waiting to collect at a baggage belt and paying additional fees is a burden that is sometimes not worth carrying. There’s always the delightful experience of never seeing the case and its contents again as it wanders off into the maze of lost objects airports accumulate. Etched into my memory, even after more years than I care to think about, is arriving at a small airport after a tortuous journey of connections and having nothing but the clothes I stood up in. On a Sunday, in 35C degrees of bright summer sun that’s not an experience I want to ever repeat. Especially with a tough meeting planned for early the next day. A free airline toothbrush was no compensation.

So, I now have a new Cabin Max Metz 20 litre RPET backpack. This is an experiment on my part. Can I live out of this tiny space for 4-days? To do so is going to require some innovative thinking. In theory, it ticks all the boxes that I was considering essential. This backpack is lightweight but offers the maximum amount of packing space given an airline’s cabin bag restriction.

The plastic material the bag is made of doesn’t feel nice, but it’s flexible and hopefully durable. The zippers look substantial and should have a long life. Now, the task is mine. How to choose exactly the necessities of life to enjoy the journey ahead. To pack as smartly as smart can be.

Air Taxi 3

Urban mobility by air, had a flurry of success in the 1970s. However, it did not end well.

Canadian Joni Mitchell is one of the most celebrated singer-songwriters and my favourite. She has tapped into the social and environmental issues that have concerned a lot of us for decades. Of her large catalogue, I can’t tell you how much I love this song[1]. The shear beauty of the lyric.

Anyway, it’s another track on the album called “Hejira” that I want to refer. When I looked it up, I found out, I was wrong. The song I want to refer to is on the 1975 album “The Hissing of Summer Lawns”. The song “Harry’s House[2]” contains the line “a helicopter lands on the Pan Am roof like a dragon fly on a tomb.” Without going into what it’s all about, the lyrical image is that flying from a city skyscraper roof was seen as glamorous and the pinnacle of success.

In 1970, prominent aviation authorities were talking about the regulatory criteria needed for the city-centre VTOL[3] aircraft of the future. Then on the afternoon of 16 May 1977, New York Airways Flight 971, a Sikorsky S-61 helicopter, crashed[4] on Pan Am’s building rooftop heliport[5]. That ghastly fatal accident reset thinking about city centre operations air transport operations.

So, what’s different 50-year on? Proposals for city centre eVTOL operations are much in the News. City planners are imagining how they integrate an airborne dimension into public transport operations. Cars, busses, trains and eVTOL aircraft may all be connected in new multimodal terminals. That’s the city transport planners’ vision for less than a decade ahead.

For one, the vehicles are radically different. Yes, the physics of flight will not change but getting airborne is quite different between a conventional large helicopter and the plethora of different eVTOL developments that are underway across the world.

Another point, and that’s why I’m writing this piece, is the shear amount of safety data that can be made available to aircraft operators. Whereas in the 1970s, a 5-parameter flight recorder was thought to be neat, now the number of digital parameters that could be collected weighs in over thousands. In the 1970s, large helicopters didn’t even have the basic recording of minimal flight data as a consideration. The complexity in the future of eVTOL will be, not how or where to get data but what to do with all the data that is streamed off the new aircraft.

Interestingly, this changes the shape of the Heinrich and Bird “safety pyramid” model[6]. Even knowing about such a safety model is a bit nerdy. That said, it’s cited by specialist in countless aviation safety presentations.

Top level events, that’s the peak of the pyramid, remain the same, but the base of the pyramid becomes much larger. The amount of safety data that could be available on operational occurrences grows dramatically. Or at least it should.

POST: Growing consideration is being given to the eVTOL ecosystem. This will mean a growing need to share data Advanced Air Mobility Portal (nasa.gov)

[1] https://youtu.be/nyj5Be5ovas

[2] A nice cover https://youtu.be/bjvYgpm–tY

[3] VTOL = Vertical Take Off and Landing.

[4] https://www.nytimes.com/1977/05/17/archives/5-killed-as-copter-on-pan-am-building-throws-rotor-blade-one-victim.html

[5] https://www.thisdayinaviation.com/16-may-1977/

[6] https://skybrary.aero/articles/heinrich-pyramid