Drone detection has become a dirty word.

A look at yacht-based drone detection strategies and why we can't just bury our heads in the sand.

In recent years, in my roles as head of superyacht security teams and on behalf of management, I have, on multiple occasions, had this passed on to me. So, I've met with companies all over the world, each pursuing a holy grail of countering this suddenly feasible threat, and inevitably obfuscating somewhat the realities of the situation.

As security personnel, we're caught in the middle, between marketing hyperbole and a blind trust in tech. We are the ones forced to juggle the expectations of our clients with the realities of physics and budgetary constraints. As a result, it's imperative we know what we're dealing with, and the key is arriving with a solid understanding of how and why drone detection works, and crucially: how and where it fails.

A relatively new and suddenly accessible threat

Even when I served in the military, just over 15 years ago, there simply wasn't the threat of commercially available drones appearing overhead with unwelcome munitions; now it's a game-changing battlefield tactic. Whilst in a civilian context, we are rarely looking at this worst case, even in its watered-down form (privacy-invasion, protest action or suchlike), it is significant enough that it absolutely must now be factored into a superyacht security strategy.

The sheer availability of small-to-medium consumer drones over the past years has now eventually kicked the regulatory bodies into action. But the task of detection isn't as straightforward as we'd like to think.

Hyperbole and the requirement for baseline understanding

Time and again, I'm learning that this topic can be as much about the reconciliation of expectations as it is about the detection measure itself.

There remain several detection technologies in active development, but by far the most common is RF-detection. In theory, this is not a difficult concept; we listen for drone communications on the standard frequencies and pick out the ones we can identify (an over-simplification, but sufficient for now). Frequently, we can deduce (if not directly decode) certain telemetry data, often including even controller locations. We present this on a sleek-looking dark-mode satellite map overlay, and everyone's happy.

... but herein lies the problem; this certainly isn't the whole picture.

Minimal viable detection

In reality, 'nuisance' drones are often not the problem; these are now heavily regulated in most airspaces, and in order to gain regulatory approval, they are very deliberately conspicuous and compliant with standards.

Our primary concern should come from those outwith: customised FPVs, detection-evading/freq-hopping, location spoofing, full or partial autonomy... Can we expect to detect these? Field demonstrations are carefully choreographed to show only selected drones (invariably the low-hanging fruit of drone detection) and almost always in very 'clean' RF environments.

In short, though, the fundamental question is the same: how do we maximise our chances of detection, avoid good money after bad, and what can we do to mitigate?

So, what should we know?

First off: whether as management, security team or captain, we must understand our threat analysis. What level of protection do our clients/guests expect?

In instances when we already have equipment installed, we must understand its capability; what type? What frequency ranges? Where is it installed? What about blind spots? What about library updating?... and critically, what will we simply NOT detect given our setup? Without a handle on this, we can't assess performance and manage our security responses accordingly.

Needless to say, not all RF-detection comes equal; whether a drone is visible or not to RF-detection depends on whether you're listening on the right frequencies and you're listening for the right things; this is to say that we rely heavily on ongoing R&D (and updates) from our suppliers. When it comes to initial spend, due diligence on the supplier/manufacturer is arguably as important as on the equipment itself.

When selecting options, relevant considerations don't stop there. Here are just a few more for pondering:

• How does it integrate into your systems (and more importantly, procedures) - if it doesn't, it's going to be ineffective in real-world use cases, regardless of how well it functions

• Is it properly adapted for moving-platform deployments? The market was built upon fixed site installations, and I have often seen these insufficiently adapted for maritime use. This includes multi-sensor deployments - think back to trig at school before we accept any claims of effective triangulation from single vessel deployments!

• Hardware does matter - developments in this field continue, particularly when it comes to 'Angle of arrival' - remember that non-compliant drones are unlikely to broadcast truthful positions; in this case, the best we can hope for is approximate direction and estimated distance.

• Incorporation of additional detection methods - very much an article in itself! But as above, remember that the further we get from the target market use-case, the quicker costs increase and the performance decreases.

What can we do?

A full discussion of mitigation is probably for another day. However, the most obvious starting point is frequently overlooked: internal response procedures.

These are founded on dynamic assessment of threat and risk (given your position, environment, guest situation, most probable risk, etc), but in all cases always rely on a timely, effective and streamlined alert/response workflow. This is why that integration alluded to earlier is so critical. A fluid and rehearsed response is always going to be what decides the outcome. Of course, this will be a situation-specific response. In many cases, alerting guests and crew will be all we have at our disposal. But this, while inconvenient, should always be the primary 'actions-on'. It is also worth considering your situation. Drone incursions at sea, or more than a few miles off the coast, are incredibly rare, and will often be initiated from a nearby vessel - this should be factored into your response plan. There is plenty more to say on response plans, so please let me know if this is of specific interest.

The TLDR

Be wary of the hype and KNOW well the gaps in your own detection strategy. We cannot simply bury our heads in the sand and blindly trust the tech; there will never be a guaranteed coverall. As much as possible, we should strive to understand the means of detection and factor in the subsequent points of failure. And with regards to mitigation, remember that timely alerting and effectively implemented responses are always going to be the first priority, regardless of what equipment you have installed.

Being aware of the underlying methods of detection is essential for weeding out unsuitable or low-value detection implementations. In many cases, even by asking the right questions, we can find it hard to get straight-up answers, so please do get in touch if I can provide further assistance on the topic!