Back seat drivers: An under valued safety net (or: On perception and decision making)

Have you ever been driving with your mother on the passenger's seat identifying all sort of potential threats? Well, this is what I am talking about: to what extent is this a nuisance and to what extent, if used effectively and appropriately, could serve us as a safety net in our safety critical work:

In human factors we best describe the way how we arrive to take decisions through what is called the Human Information Processing Model.

In a nutshell we receive data from outside in the form of stimuli. This 'raw data' is received through our senses. We perceive this data through an internal processing system (something like when a computer processes data to produce information). Internally we process the data through a series of iterations between our short term (or working) memory, which temporarily keeps the data we receive and the longer term memory from which we retrieved learned information (that we acquired through previous experiences). We mish-mash, reiterate, and make sense of the situation (situational awareness). This awareness is kept in the short term memory, and through it we take decisions.

I put a drawing to illustrate the process.

[caption id="" align="aligncenter" width="272"] Human information process model[/caption]

In safety critical operations (but not just) we want to take correct decisions, and using the model above, the correctness depends on a number of factors which include our ability of physically receive the data, to have a will functioning long and short term memory and that the long term memory contains relevant content (skill / knowledge / experience).

But we also depend on perception.

And perception can be tricked. What we perceive depends on context, on the situation, on our long term memory and on a number of other factors that are beyond our control. Beyond our control because we have evolved this way. We need this perception processing to help us make sense of the world, it is just that sometimes, our perception is not adapted to the situation and it tricks us...and we take wrong decisions...

Watch this very interesting TED video on perception if you have any doubts on what I am saying.

 Convinced?

Well, then what do we do?

The first step is to be aware of it (as many people unfortunately are not and think that what they perceive is the same as what is physically out there.)

The second is to find methods for mitigation and what can these be in our safety critical operations?

Well here is some advise:

1. If we are designers of systems, procedures, programmes, etc. we should be aware and sensible about perception and how our brain works. In small projects where resources and impact are limited, we should consider whether we could create situations, where without wanting, the minds of those operating could be 'tricked' into wrong decisions. If we are talking about bigger projects, then we should involve Human Factors Experts and have a process like Eurocontrol's Human Factors' Case.

2. If we are those on the hot seat, we will need to find ways to mitigate. Here are some:

Train ourselves to identify patterns where things may start going into a wrong direction: Example of these, is to be aware of the environment and identify feedback received by it, and start noticing and questioning when this feedback seems 'strange'. It is like sharpening our instincts. This can be done through training, but mainly involves experience, power of observation and self criticism.

Another mitigation is team situational awareness. Having a team working we are adding more resources that can identify patterns and we are adding diversity to perception bias. Team situational awareness means that other team members, are aware of what you are doing and are following the same situation but from a slightly different role / angle. The trick is not to do the same tasks, as here we are reducing diversity. It is to have complimentary tasks whilst sharing similar awareness of the situation. How do this? To start with we need to have a minimum degree of competence in the tasks. We also need to be familiar with the team, and with the individuals composing the team, how we all work, procedures, practices, etc. This is a process which is built over time and through meetings, sharing of experiences, etc. We need to have agreed ways of working as a team, not to give each other unnecessary frights (remember your mother on the passengers seat giving you the fright of your life?) Then the person making the decisions (I am thinking short term decisions, such as taking actions whilst controlling aircraft or maintaining a system) should articulate the decision they are about to take. This allows the other team member(s) to follow the process and to be aware of what is going on. Other team members would need to keep to their tasks and not get 100% involved in the task of the team member they are helping out (we're not talking of 2 drivers here). For example, if we have a situation with an executive controller and a coordinator, those who have experienced this would agree that whilst both following the same situation, if the roles are respected, the coordinator is able to keep a view of the broader picture, whilst following what the executive is doing. On the other hand, the executive is focussing on responding and taking an active role in the decisions. As those who have observed someone in the hot seat would tell you, from the back seat you see things that you wonder how come the person acting has not yet seen, yet when the observer switches seats and goes into the hot zone, his / her perspective also changes. This happens in every day life with back seat driving - thus the example which gave the title to this post. In ATC this happens with the instructor role for example.

What we need to do is to make the best of this situation and use it to identify issues that are being missed by the hot seat and assist in this way. This is team situational awareness...

To summarise:

  • We experience the environment in a processed manner called perception.
  • While what we perceive feeds into the decisions we take, what we perceive is not necessarly what is going on. This is not a weakness, it is a state of us being human and alive.
  • In order to mitigate this danger we need to first be aware of this.
  • Then we can do a number of things to mitigate further. Two of these are:
    • Train ourselves in identifying feedback patterns from the environment where things seem to be wrong, a sort of sharpening our instincts. This comes with training but mostly with experience
    • Build team situational awareness. For this we need a team. We need to share tasks in a complimentary manner to allow things to be perceived through at least slightly different contexts. We need to articulate our decisions, and we should not be afraid of having situations where one of us is on the hot seat and the other on the back seat. Both are complimentary functions.

 These are the kind of things we at AriaTM teach and facilitate during our Team Resources Management sessions.

Big Data and Air Traffic Management

According to webopedia (webopedia.com) Big Data is a buzzword, or catch-phrase, used to describe a massive volume of both structured and unstructured data that is so large that it's difficult to process using traditional database and software techniques.

[caption id="" align="aligncenter" width="571"] courtesy of colocarionamerica.com[/caption]

In fact I am using it exactly as that: As a buzzword in my title to try to catch your attention.

Today I have attended a very interesting workshop about Data Science in aviation organized by Innaxis (innaxis.org). But what is data science? Ï have checked a few definitions, and the one I like best is the one given by searchcio.techtarget.com which says that: "Data science is the study of where information comes from, what it represents and how it can be turned into a valuable resource…"

So, what about data science and air traffic management?

Well, if we think about it, increased awareness of data science (and the use of buzzwords like ‘big data’) in the last few years is one of the next (or present) natural evolutions of the greater digital revolution: first we accelerate the generation of data, and commodatise its storage and processing power, and next we think of generating value from all this data.

But indeed it makes sense. To take other industries as a benchmark, companies are aware nowadays not only that the data they have collected and are collaterally collecting can have value to generate new business and to render current operations more efficient, but they now have started to pro-actively look for ways of capturing even more data so as to accelerate further this value creation - Thinking about how companies like google, facebook, credit card and telecommunication companies, just to mention a few, are gathering our data on purpose (and for free) to generate further revenue for themselves is scary and somewhat perverse... However, it is reality and the aim of this workshop was to see how all the data that we are gathering in aviation (accidentally or purposefully (structured or unstructured, if I understand the jargon well)) can be used to generate more value for us.

As one of the presenters put it, the value could be one of three types:

  1. to generate new income,
  2. to help in decision making and to render the system more efficient, and
  3. to reduce risk, thus making it more safe.

And the three objectives are applicable for aviation: In ATM, we are generally looking at making the system safer and more efficient. Other segments of aviation are looking for new income.

During the workshop we listened to 4 different operationally-concerned presenters (2 from the airline world, another from an aircraft manufacturer and one from ATM) explain that they are collecting a lot of data and that some of it is being used to create value, for example to create capacity, to compute the best flying profiles or to improve the airline safety records. Yet, I felt that the underlying thread was one which said: we have a lot of data which we are under using: Data Scientists, please come help us find ways of how we can generate more benefit from this data. (and here we are back to the big data definition above…)

The workshop continued then with a series of other presentations, this time from professionals in the field of data science whose objectives was to educate us and to give us more background information and to inspire us.

I am sure that the day was fruitful for many, as the idea of gathering aviation operations and data scientists in a room will give many of us ideas for the future.

As for me, apart from thanking once more Innaxis for making this workshop a reality, I come home convinced that data science is a necessity in ATM, if we want to move ahead, if we want to better understand our complex ATM system and if we want to be wise about tomorrow’s decisions on how to enhance safety and efficiency in our industry.

It also gave me an appetite for those mathematical formulas I left 19 years ago; but that is already another story…

...(by the way, can anyone predict the next evolution within the digital revolution? Working on this from now will be worth a rich gold mine...Anyone? Data scientist? Let me know ;)

Another (most probably discriminating) way to reduce fuel consumption in flight

Coincidentally, in the last post about deconflicted 4-d trajectories, I wrote amongst other things on what seems to me to be an innovative way to see efficiency in terms of time elapsed and fuel consumed (rather than the classical distance covered). [caption id="" align="aligncenter" width="228"] Jeh Wadia, the Managing Director of GoAir, stands amidst women, dressed as flight attendants. Reuters[/caption] I write coincidentally because today I read about yet another way of consuming less fuel (and consequentially rendering your flight more efficient...). This idea apparently goes to the management of GoAir, an Indian low-cost airline. The guardian [click here to read the article] reports that management of GoAir decided to start hiring female only cabin crew. This, they say (and its true but discriminatory in many countries - don't know about India) will reduce the flying weight by 20kg per cabin crew member, making savings for the company of over €380 000 per year. Well way to go GoAir...maybe next would be to cap the cabin crew's maximum weight, and then, what will be next? If you read so far and are interested in more, here is a better article

De-Conflicted and efficient 4D Trajectories through a Centralised Service

Recently I have had a very interesting discussion about the future of air traffic management with a person who has for a long time studied future concepts to improve efficiency in Air Traffic Control.

4D trajectory: Image embedded from www.onera.fr

I was in Bucharest on a 2 day trip to give presentations about ATM and ATC from an operational perspective to Aeronautical Engineering students at the Polytechnic University of Bucharest. The vice dean, who has read a PHD in 4D Trajectories Optimisation started a discussion on what the vision may be regarding 4D trajectories and on the computing of the most efficient trajectories for flights in terms of energy consumption and time.

He said that whereas he agreed with the common concept that the key towards further efficiencies in the above mentioned areas is to be able to push the traffic situation into proactive management of already de-conflicted situations (eventual gate-to-gate precision trajectory clearances) rather than legacy reactive / tactical ATC based achievement of separation, there were two elements that needed to be enforced. These were:

  • The idea that the trajectory of flights which are at once both de-conflicted from other flights and also the most efficient, should be entrusted to a centralised ground system, and
  • Secondly, that the main criteria for the most efficient trajectory was one based on time and fuel consumption and not on distance covered (that the most efficient route might not be the shortest one).

1. Ground based centralised de-conflicted and efficient precise trajectory clearance:

The idea makes sense, the system should be based on a centralised processor that analyses all flight requests and parameters stemming from all ANSPs and proposes the best trajectory to flights. The term best would be defined as a trajectory which would be de-conflicted from all other flights and most efficient. ATC would be then responsible for monitoring the flight and to act on last minute changes due to unforeseen circumstances. The efficient trajectory would be based on meteorological reports and forecasts, thus making the best use of prevailing winds, etc.

In this model, the Reference Business Trajectory for a flight would be prepared by a centralised regional service based on information from various stakeholders including ANSPs and weather services. The RBT would then be accepted by the airspace user who would fly it knowing it has a high degree of stability (it would not be likely to suffer major revisions whilst in flight), and being the most efficient.

This idea to me makes sense as it marks an evolution of the direction the management of the ATM network has been taking, from Flow Control, to Network Management to perhaps Enhanced Network Management. History has shown that a centralised responsible body, pushed and governed by its stakeholders (such as the case of CFMU and NM) are the most efficient responses to unleashing and enforcing the necessary situations to improve flight efficiency.

2. It is about time and fuel consumption and not distance

 Aircraft with sails. Image courtesy of blog.flying-hobby.net

The second aspect that I found quite innovatory in this discussion I had is that efficiency should be seen in terms of flying time and in terms of energy consumption and that this does not always correlate with distance covered.

The way things are presented today is that the closest we get to the great circle in terms of distance covered between two airports, the more efficient our flight is. But this may not be true if to fly the great circle one ignores wind, for example. (If a head wind component is more important for a given flight, should it fly a great circle trajectory than it would be if the flight steers away from this great circle trajectory into more favourable wind and therefore saves more fuel and/or arrives earlier to destination by doing so – a bit like the case of a boat that has to choose between navigating against the current for a direct route or to let the current steer it in terms of maritime navigation. We need to become more sophisticated when talking about flight trajectory efficiency and how we calculate it. Is it the distance covered? (as we see it shouldn’t be), is it the flying time? Or the energy used to fly? (possibly it should be a calculation based on the component of the last 2).

The above for me is very simple to understand, yet it seems to me as a break through as we always tend to talk about trajectory inefficiencies based on the deviation from the shortest (distance) route and then by computing the fuel and/or time lost as a direct conversion of the distance, whereas what we need to be talking about is on how to make trajectories to be agreed based on the knowledge of winds and of engine performances.

The above seems to me more feasible to be achieved using a centralised service rather than a scattered one. In terms of applicability in the near future I see also that in terms of phases of flight, it may still be difficult to manage these type of trajectories in terminal airspaces and in and around airports, because the incognita in these areas seem to me still too complex to compute and manage (from passengers arriving late to the gate to GA traffic making unpredicted manoeuvres in the air and airside vehicles on the ground.)  However, I could imagine a centralised service presenting on behalf of ANSPSa trajectory contract of this type to the flight in question, based on the current situation and de-conflicted from other traffic from the moment the flight leaves the terminal area of departure to the point it arrives to the terminal are of arrival as something that if worked upon could be made feasible in a foreseeable future (This close to Sesar’s Operational Improvement Step:” Use of Free Routing from Terminal Area Operations-exit to Terminal Area Operations-entry” but rather than ‘free routing’ we should be aiming for a de-conflicted and optimum-efficient precision trajectory clearance from Terminal Area Operations-exit to Terminal Area Operations-entry)

I think this, based on a centralised service and on a real time system wide information management exchange and on the concepts of de-conflicted precision trajectory clearances  which are optimum efficient in terms of a computation of time and fuel consumption, should at least make the base for a serious study into feasibility and into a shared operational concept...

Such was my discussion in Bucharest. I went to teach and as often happens, I came back feeling I have learnt something very significant about the future of air traffic management!

Training offered by AriaTM

Since a few months I have developed my own training courses under AriaTM revolving around the improvement of safety and efficiency in ATC through the better preparation and training of ATS personnel.

I have developed the training plans and training material for
  • Team Resources Management
  • On the Job Instructor Trainingariatm_just_logo.jpg
  • ATS Supervisor Management
  • Competence Assessors and Evaluators                         
  • Classroom Instruction Techniques
  • Simulator Instructor Training (and all the refresher training for the courses above)

The courses can be delivered to different Air Traffic Management personnel including ATCOs, ATSEPs, Apron Controllers, Airport airside related personnel, etc. The material is developed based on guidelines and specifications when these are available and on best practice and personal experience, either giving the courses, listening to attendants or from direct experience.

I normally collaborate with other professionals both in the design and in the delivery of these courses, which, I believe, makes the content even richer. I have a 2d portable simulator that I use to enable role play in courses like OJTI, Sim Instructor or Competence Assessor.

My principal model is to go to the client and deliver the training reducing to a maximum the administrative overhead and concentrating on the quality.

I am also open to deliver these courses to individuals if so requested.

Presently I am also working on support to improve English language proficiency.

Should you want more information, please do not hesitate to drop me a line at max@ariatm.com

ATC open day

A few weeks ago I attended an ATC open day.

It was a very interesting day which had the main objective of introducing Air Traffic Control to the public. The day was divided into different parts ranging from presentations about air traffic control, to demonstrations about controllong aircraft in a simulator to finally a hands on for those who wanted to try ATC. The presentations and the demonstration were given by air traffic controllers but the event was organised by a non profit organisation made up of non professionals who are enthusiastic and passionate about air traffic control. This event led me essentially to two thoughts; the first being, how beneficial it is to organise similar open days or awareness days for the public of a profession which otherwise stays in the shadows most of the time (and one normally hears about it negatively: either because of an incident, a delay (the pilot saying it is due to air traffic control) or because of a strike. The second thought was: had it to be a group of non professionals to have this great idea and organise this day? Where are the professionals? Should we not do more of this in every corner of the world? Perhaps finally there will be more who will understand what the work is about, a few more who wish to become air traffic controllers, a few more who move away from the cliche of "oh it must be a stressful job" and closer to understanding a bit more what this stress is about, what safety is about, what controllers worry about... I think it would very much benefit the profession, who needs to keep in contact with the public. So if you are part of a professional guild of Air Traffic Controllers, I encourage you to organise a similar yearly event in your community to preach about your profession. Also coincidentally last week the IFATCA general conference took place. (I hope it was a great success!) I suggest to IFATCA to consider to run an ATC open day to the public in parallel with the general conference and with the regional conferences. When the conference goes to a place, apart from lobbying the local authorities, we could reach the public there and educate them about our profession. Finally, if you are in Spain and are enthousiastic about air traffic control I suggest you join the non profit organisation who organised this. They are called OECAV (www.oecav.org). (not linked to IVAO) They are very proactive and doing a good job in our name. Supporting them is the least we can do!

Security breach: Remotely hacking into aircraft FMS

I do not consider myself someone who alarms himself easily, nor someone who likes to alarm others. However, recently, I came across a presentation, which was also was reported on some websites, about someone, Hugo Teso, who reported having managed to remotely hack into the FMS system of an aircraft by using ACARS. This report was done during a security conference in Amsterdam earlier this year.

He reported he managed to hack into a replica of a Flight Management System (FMS) in a simulator and cause the aircraft to crash. If this is the case, the implications may be great and security upgrades should be implemented fast on the vulnerable systems. In the presentation it says that he used two vulnerable systems which are in wide use: he used ADS-B to find an aircraft, as the ADS-B gives information about the aircraft's 4d position (time, lat, long & altitude), then through ACARS he contacts the aircraft and tries to hack into the FMS (as ACARS has a connection to uplink information to the FMS, so the connection exists).

Once into the FMS he was able to manipulate it.

Well for me this is very worrying, much more worrying than bringing liquids which have not been verified on-board.

The only comfort is that the report is proactive (it is done before something has really happened) and that it concludes with a note that work is being done by EASA to solve the security vulnerabilities.

Aircraft Altimeter Failure

"An altimeter failure on a PC12 aircraft caused a 2000 ft discrepancy between the displayed altitudes on the aircraft’s two altimeters. This contributed to a very serious airprox which identified a number of safety issues of interest to the aviation community...." This is a very interesting read from Skybrary...to read the full article click here