Tearing up the rule book
We found out on the 11th November that the DfT has made a decision to ignore the EASA Open Category C marking system. The current expectation is that they will issue an amendment to the Implementing Regulations before the end of December with the changes coming into force on 1st January 2023.
In many ways this is unfortunate as it is a system which was about to bloom with DJI’s November release of the Mavic 3 Classic. This drone was predicted (not as far as the name, but in terms of a Mavic 3 variant being the first true C1 launch) in the blog A Guide to NOT Buying the DJI Mini 3 Pro – Eyeup Aerial Solutions. It is therefore unfortunate that the DfT/CAA have waited until a drone has come onto the market that can be legally used in the A1 sub-category.
However, the announcement does provide the UK with an opportunity to change its approach to the regulation of Open Category drones. This blog is not intended as another prediction, but it is Eyeup’s idea of the direction some of those decisions may take us and, most importantly, why.
Note that the DfT doesn’t have to take until December 2025 to come up with a new set of rules. They can do it as soon as they are able…I’m just here to try to help the debate and move things along.
Let’s get Specific
It also refuses to confine itself to the Open Category. There are some important elements to the current C class model that lend themselves to use by professionals under an Operational Authorisation in the Specific Category. The CAA has the ability to change what is allowed under the Specific category using a system of pre-determined risk assessments. These proposals seek to open up some of the advantages of the previous Open category regulations to Operational Authorisation holders, whilst retaining the oversight that the CAA has always required of professional operators.
I make no apology that the changes outlined below may seem to favour “commercial” operators. That’s where I come from and is the sector I’m primarily interested in. I leave it to the hobby sector to understand the regulations and come up with their own reasoned ideas. Let’s take this opportunity to have a debate and show the world that UK users are a responsible community.
You will find more of this debate going on over at the Geeksvana YouTube channel as the CAA and DfT hopefully enter a period of debate and consultation on how the new rules should be shaped.
Show us the money
Let’s get this one out of the way. I believe there needs to be a re-think around the concept of the basis of a flight. The removal of flying as a commercial activity from the “Risk Register” has, in many people’s eyes, been a retrograde step.
Yes, it “opens the market” through de-regulation, but at the same time it opens the door to reduced quality of output through the availability and use of small-sensor drones in the hands of novices out to earn pocket money. While this is great for those earning small sums for lower-end jobs, it also starts to nibble at income of professional pilots at the lower end of the market. There is nothing wrong with fair competition, but in this case the market all barriers to entry have been completely removed at a level which used to help sustain new entrants to the market and help them develop their skills. The market for serious entrants who have invested in training and equipment looks like a broken ladder, with the lower rungs (low-end jobs), broken or removed. This is not sustainable.
The UK unmanned aircraft industry is far from mature and there are few openings for new pilots to join established companies and get themselves mentored over a training period. They generally fight their way up through the market from easier, “low-end” jobs of relatively low complexity and gradually learn their trade. And the barriers weren’t purely financial. A practical flight test, necessary to gain the old NQE recommendation (equivalent of a GVC), meant that the remote pilot required a modicum of flying ability. Annual renewal of their permission meant required at least a basic level of commitment to knowing about regulatory changes.
When this requirement for a “commercial licence” existed the CAA used to insist that it was from the point of view of safety. “If you are flying for a client”, they would patiently explain, “then there are additional pressures to perform and capture the data in situations where perhaps conditions are less than ideal. This can lead to human factor errors and an “incident”. Professional operators know only too well that this can be a very real risk whether on set or on a construction site.
Then it was decided that “commercial” differentiation would not be recognised within the regulation and it was amazing how contacts at the CAA did a U-turn that made Liz Truss look like she was stuck on rails! Suddenly, these commercial drivers were no longer a risk factor at all. Everything is now purely about the weight of the drone and where it was being flown. Whilst the CAA will agree that human factors are the major driver behind most aviation incidents, apparently the whole “client pressure” risk has entirely disappeared. Who’d have thought?
My belief (and that of many others), is that the commercial differentiation needs to return. There are few areas of transport that do not require some form of licencing to earn money, and it is strange that drones are singled out in this regard, given an identified level of additional flight risk.
Drone Rules: REQUIREMENTS FOR FLYING IN THE OPEN CATEGORY
WARNING: THE FOLLOWING SHEET IS NOT LAW AND IS NOT THE CURRENT CAP2012-BASED DOCUMENT. IT IS AN EYEUP AERIAL SOLUTIONS PROPOSED MODEL TO REPLACE THE CURRENT VERSION FROM 2026.
Categories, Sub-categories and Sub-sub-categories!
This proposal assumes that we will retain the concept of different categories of flight. As it stands, the “Open” category is light touch from a regulatory point of view. Buy a drone, register, get the appropriate competence (qualification) and go fly. There is also a “Certified” category which will cover operations which are considered such high risk that very special controls are required, such as air worthiness tests on the unmanned aircraft. Think heavyweight BVLOS delivery services or air taxis and you’ll get the idea.
In between these two is the “Specific” category. This covers operations that are unable to follow all the rules of the Open category. Heavier sub-25kg drones in congested areas or dropping of articles would be a couple of examples.
The major regulatory difference between the Open and Specific categories is that the CAA maintains oversight of the Specific category operations. The Operators in this category have to operate against a “specific” set of rules written into an Operations Manual and this manual is submitted each year for review by the regulator. It isn’t easy and many operators in this space feel that the oversight is onerous for a reducing benefit. Why is the benefit reducing? Well, now an open category flyer can come and sweep up some of the lower value bread and butter jobs with few of the regulatory overheads born by the Specific category operators. We have to ask if this is fair to those who have invested heavily in training, qualification and technology.
If you’ve clicked on the Eyeup version of CAP2012 above, you will notice that the Specific Category has sneaked its way into what used to be an Open category only crib sheet. This is because if we see risk as a spectrum, then we need to reflect the blending of the Open category into the Specific category. Apart from anything else, it indicates that there is a level above the Open category and the competence level required to operate there.
Let’s dig into the Open sub-categories
A1 is a confusing sub-category from day one. Why? Because under the EASA regs it is actually two sub-categories in one, with different flight rules. It includes “flying over” uninvolved people and “no intentional flight over” uninvolved people. This is overly complex, easy to misunderstand and therefore, not as safe as it could be.
Flight “over people”
I am a fan of the sub-250g “fly over people” sub-sub-category. I believe that for anybody wanting to fly for sport or recreation can get all they need from a drone of this weight. The sensors are getting larger and the capabilities in this class will continue to improve, particularly as the huge US market has a “no registration” benefit attached to these tiny machines. Because of the weight of these drones, the risk is inherently low. Furthermore, with appropriate training (let’s call it the A1CofC), I see no reason why this “flight over people” couldn’t be extended to the higher ”transitional” weight of 500g.
Why is this? I’m afraid I’m going to diverge…
The term “intentional overflight” is essentially meaningless and I believe would be unenforceable except in some very rare circumstances. The reason is that if a drone is moving and is at any meaningful height then the laws of physics mean that even in the event of total failure, it will never hit an item directly below it. Momentum will continue to carry it forward and it will land a distance away from whatever was below it when the failure occurred. Although you overflew a person, they were safe.
On the other hand, you may be flying on a route that is perfectly legal, but towards an uninvolved person (or the person may appear from a doorway for instance in front of the flight path of the drone). If the drone fails now, its momentum will carry it into the person. There was no overflight, no intention of overflight but somebody had a drone hit them.
These two examples make it clear to me that in the incredibly rare case of a mid-air drone failure, the concept of intentional overflight is not useful. If the DfT wants to mitigate against the risks of hitting uninvolved people, then they have one way to do it and that is to keep drones away from areas where there are people. They have chosen not to do this so should really accept the physical reality of the consequences of that decision.
Flight close to people
A1 Sub-Category <250g
I do not believe that hobby fliers really need to be able to carry the sort of sensors now being built into the new sub 900g class of drone envisaged under the EASA regulations (previously recognised as “C1”).
I would therefore limit sport and recreational users who (it should be remembered, may be flying entirely uninsured), to these sub-250g drones.
However, the past two years has illustrated through social media, an appalling level of understanding of the regulations amongst sub-250g drone flyers. Many do not bother taking the very basic online CAA DMARES Flyer ID test. We are therefore left with a ridiculous scenario where a group of operators are expected to follow the rules with absolutely no requirement to know and understand those rules. Many of them don’t even understand the basic qualification of reading the user manual. This situation shouldn’t be allowed to remain and the DMARES test should be the minimum standard of competence for any weight or registered drone.
A1 Sub-Category <500g
I would go further than the current regulations and allow overflight of uninvolved people under the A1 sub-category by drones <500g. As per the current regulations in terms of competence, this would either be covered by a “legacy” A2CofC, or by a newly introduced A1CofC.
What happens to the sub-900g category of drone? In my mind, these are potentially more dangerous in untrained hands. I would suggest that the class weight be increased to 1kg (allowing the attachment of additional safety accessories such as prop guards) but the use of these drones be limited to 785/2004 insured flyers only who proven practical flight capability.
The reason for this is that when things go wrong with drones, they can go very wrong, very quickly. The bigger the drone and the more complex the flight environment, the more likely things are to go wrong and the worse the consequences. It strikes me that the minimal DMARES Flyer ID test in the current regulations was never a sufficient level of competence to allow a complete novice to take a drone of this size amongst the general public.
The A1 sub-category would be limited to sub-250g drones with an upper speed limit of 19m/s, pretty much as it is now for flight over people. But I would remove the sub-900g “C1” type drone and the confusing (and impossible to define) “No intentional flight over uninvolved persons” sub-sub-category altogether.
What to do with this sub-category? It has always seemed an odd one to me. The training piece has been reasonable, though I would have always argued that the A2CofC training should have been the minimum for A1 class drones.
The 1:1 rule rapidly makes the use of these drones in congested areas a real pig to manage. Because it is a horizontal clearance, the moment you are up to 30m height, ground risk management soon becomes impractical. Imagine any sort of town centre job where, strictly speaking, you are having to control several streets around your drone while carrying out a roof survey. As soon as you look at flying across an area for cinematic shots you may as well give up!
These drones are going to be expensive as well. For a C1 drone we know you are in for £1800-£2000 to invest in a decent drone eco-system (platform, filters, batteries, chargers etc). The smallest DJI drone that may achieve a C2 certificate would be the Mavic 3 Enterprise £3250 to £6700. I can’t begin to imagine what a full mass (up to 4kg) C2 drone will cost…but it isn’t general hobby money.
If you are investing at this level, I would humbly suggest that you are probably working commercially and would benefit from working under the Specific Category controls.
I would therefore either scrap the concept of A2 drones in the Open Category and do what I’ve done with the “C1” drones and drop them into a revised Specific category authorisation. In my mind, the use of these heavier drones in congested areas requires a higher level of training than the current PDRA01 authorisation allows. This thought process is full aligned with the existing OSC-based authorisations where multiple hoops have to be jumped through to obtain a 5m take-off and landing permission. Although the newer drones are safer, as I’ve stated above, when things do go wrong, they can go wrong quickly and badly. To allow novice pilots into heavily congested areas with 4kg drones and a 5m take-off clearance, with no additional mitigations seems to be the height of foolishness.
I would therefore propose an additional Pre-Determined Risk Assessment with an appropriate training syllabus and practical test and operations manual update. Let’s call this PDRA0X.
Specific Category Bonus
But, under Eyeup’s proposals, you wouldn’t need additional training to be able to fly closer to people with a drone heavier than 250g.
Do you recall how I cut the “C1” sub-900g drone out of the Open A1 sub-category altogether? Well, here’s a nice surprise. I think that these lower mass drones can be flown perfectly safely at lower clearances provided this is being done within a controlled system with CAA oversight and by operators with financial “skin in the game”.
I would argue that sub 1kg drones could be used by OA-holders (while flying insured) with the significantly reduced clearances envisaged under the current “C1” arrangements. It may be necessary to introduce guidance on a minimum height for overflight (to allow crew reaction time in emergencies), but the ability to take off and land at tiny clearances can be managed effectively under an operations manual.
Increasing the take-off weight from 900g to 1kg or higher may allow for additional safety equipment such as lighting or prop guards (making take-off and landing even safer for uninvolved people) to be attached to an EASA C1 platform. It also allows us to steal a march on our European cousins by allowing enterprise versions of platforms like the Mavic 3 Enterprise Thermal (930g), to be used at reduced clearances.
But what are the “standards” for the drones?
We now circle back to a key question. I have mentioned categories above and used weights pretty much in line with the current EASA “C” classes. So how will drones be standardised in this brave new world?
Do you know what…I don’t think it’s that difficult A combination of market forces and existing standard from the rest of the world (and yes, I include EASA in that statement) mean that drones will be safe. Let’s face it, they are already when you look at all the statistics. We also know that even if a dud product makes it to market that it is extremely unlikely that people will be hurt. We also know that such products are whipped off the market in short order and either the issues are resolved or the product dies. Has anybody tried buying a Go-Pro Karma recently?
So, as a country, let’s do the smart thing and ride on the back of market forces. we will be using equipment that will already be being designed and built to very high standards for other, similar markets and can therefore simplify our own standardisation to a few simple points. These may include weight, maximum speed (=maximum energy) and sound levels…but the key thing is that they can be simple…because the work has already been done elsewhere.
Get on the training train
Finally, I will touch on training.
To my mind either drones present a real risk to aviation and people, or they don’t. It is clear to me as a seasoned operator that things can go wrong (AAIB reports provide evidence of this) and that in many cases human factors are to blame. That means that we DO need training, but it also tells us that the training may receive is not up to the job.
In Eyeup’s ideal world, training will be strengthened at every level. A FlyerID should be the minimum level of competence for anybody flying a drone big enough or equipped such that the operator requires registration. This is only the equivalent of being run through the basic safety briefing at a model aircraft club…which was the old way of getting into unmanned aviation. It is not onerous, can’t be failed (due to multiple retakes) and acts as an educational process. But hey, let’s make it a little more relevant.
An A1CofC to replace the current A2CofC (see arguments above) is probably fine at the level it currently sits. I would prefer more evidence of flight competence and given the number of PDRA01-holders out there, would try to find an easy way to allow these operators to put these newbies through a series of defined tests. I can almost smell an app being prepared to make this a simple efficient transaction.
The next level up would be the GVC training. Come on CAA, let’s get this under control. Eyeup has found itself explaining the regulations to quite a few managers of RAEs. Yet these organisations are training others! Don’t get me wrong, I don’t mind the work, but I can’t help thinking there is an oversight deficit here.
In terms of the training itself, we know from questions asked in forums that many students are left without a thorough understanding, either of the regulations or where to go to find the answers. I believe that the lack of post-sales support is as a direct consequence of the race to the bottom in pricing. This in turn has been supported by a COVID inspired allowance to avoid face to face training and do everything by video. Not training face to face by video…no, simple play the video and follow it when you’ve got time. I’m sorry, but this isn’t training in my book.
The interplay between students in a classroom (whether real or virtual) is where a lot of learning is done. Questions are asked and answered and point that are confusing can be covered in a number of ways. But hey CAA, if you are reading this, I hope I shouldn’t have to explain the obvious to you.
Finally, we come to the practical flight test. At the moment, an operator can turn up with a sub 250g drone that has GNSS systems by default and pass the test. They can then buy or hire a Matrice M200 and go and do a job in a complex city environment. Those environments can occasionally play havoc with GNSS systems and you can find yourself in ATTI mode. We know from AAIB reports that this has happened and that unwary operators have been caught out without ATTI flight skills. The CAA even drive OSC holder to practice ATTI mode flying, as though it is OSC-holder causing the problem. Meanwhile, new pilots continue to be fed into the industry with potentially zero flight skills should the worst happen and satellite lock be lost.
Come on CAA…do the right thing. ATTI mode flying must be made compulsory for the GVC practical flight tests. If nothing else, it’ll provide a market for all those Phantom 4 Pros sold so that operators could jump into the C1 category drones.
And so, we come full circle…