Anti-Drone Systems? Assessing the need with Alex Roskoss, COPTRZ Technical Manager
The Future of Drone Detection
Drones will play a massive part of our future, in more ways than we today can imagine. They are going to make changes to how we do things, changes on a scale we have not seen since horses were outpaced by motorcars. Much as back then, applications which were not conceived of, even by the early pioneers of the horse-less carriages, will come to the fore and change the way we do things forever. By 2030 up to 1.89% of UK GDP will be drone related from zero 20 years ago. As there are with computers or vehicles, there will be operators in all sectors of the economy from governmental and commercial to leisure and personal users. Across these sectors there will be good safe competent actors who follow legislation and regulations and there will also be foolish, misled and malicious actors who present risks to public safety. A UK survey shows 1 in 7 drone users admit to operating their drones outside the legal limits despite more than 97% knowing the rules.
The Threat From Drones
We have already witnessed safety and security issues with the few drones which are being flown now. Systems designed to effectively control ground access to sensitive areas are simply being flown over without any effort from the transgressor and often with no awareness at all from the transgressed. Prisons have seen drugs, phones and contraband being trafficked in by drone with up to 40 flights a month at a single facility. This is a clear case where traditional effective high security measures have been shown as outdated by low complexity drone technologies with low skilled operators.
Moving contraband joins nuisance to shipping and airfield incursion risks as reported drone security and safety incidents already seen in the UK. Accessing sensitive personal data from air-gapped networks by drone payloads and camera invasion of personal privacy have been reported outside the UK. We have seen crude munitions being carried by simple consumer drones in conflict zones with deadly effect. Drones carrying electronic distruptors, chemical, biological and radiological payloads are all very conceivable in the coming years with the technologies needed already widely available.
With the good the bad and the incompetent all flying drones now and in increasing numbers in future, what are appropriate changes to make? On a governmental level, rules are being revised, registration systems being used and improved. These measures will regulate those abiding by the legislation and allow more ready identification of those operating outside of it. However unlike motor vehicles, watercraft or aircraft, you cannot easily see a drone being transported and it can be even harder so see one being flown discreetly. Progress in controlling the regulated drone airspace is increasing but looking here for all the solutions is overly optimistic. Drones now in circulation don’t meet the proposed regulated airspace standards so controlling these will need the consent of the owners to upgrade or retire these units. Fully and heavily automated drone flights are coming soon to controlled drone airspace, here again errors in systems, configuration mistakes and malicious actors could all see drone risks.
Assessing Drone Risks
Any entity with privacy and security concerns or with safety obligations to their staff and the public should be undertaking risk assessments to know the appropriate response to the multiple threats currently posed by drones. All of these assessments need at a minimum to include an appreciation of how many drones they are exposed to, what types of drones are these, what are the drones flight activities and who is operating or overseeing these drones above or about areas of interest to the entity. Without some idea of this basic data, any drone risk assessment may be flawed. If there are frequently drones performing automated flights to a neighbouring property carrying potentially dangerous goods, this poses a different risk from occasional leisure flights by small drones to sightsee a natural feature hundreds of meters from the site.
Once you know the relevant information about drone activity around areas of interest you can create a reality based risk assessment about drone risks. The type, activity and number of drones and their likely payloads will influence the risk assessments. If protection of physical or human assets is the dominant element of your risk then numbers and sizes of drones in your proximity will be the key influence and measures to protect your people or plant against drones could be suitable mitigation. For assessments where information protection and physical security are more important the key elements of appropriate protections could be methods to exclude all drone systems or harden the vulnerabilities.
To gather information about drone activity and get adequately informed as to the risks there are many drone detection, tracking and identification systems available. All of these provide some awareness of the local drone activity however the specific detection mechanisms used and the quality of the information captured vary and these can have a big impact on the detection, tracking and identification information achievable.
Anti-Drone System Performance
The ideal performance requirements for any detection system are to be sensitive, specific and precise. High Sensitivity is desirable so as many drones as possible are detected, this typically means a wide range of drones across a wide range of distances can be detected. A system should be highly Specific so there are few non-drones reported as drones and you can operate with fewer interactions and operator time spent on non-drone events. Good Precision is important so the detection events can be discriminated by number of drones, locations, altitudes and drone types. A combination of all three capabilities provides the best detection. Without any one of these the detection system is seriously compromised.
Drone Detection Technologies
The three leading drone detection technologies are vision systems, radar systems and radio frequency (RF) systems. Each has its merits and strengths and each also has its weaknesses. I will briefly address each of these three technologies here but this is not an exhaustive list of drone detection techniques. You can find out more about these technolgies by downloading our beginners guide to airspace security.
The vision systems use camera systems and automated machine-learning monitoring to identify drones in these camera signals. Vision systems clearly work on line of sight from the sensors to the drone. To achieve 360° coverage multiple high resolution cameras are needed and the sensitivity of the system is dependent on the resolution of these cameras. These need powerful image processing systems to process the massive amount of information incoming and detect drones in these videos. To simplify this a single higher resolution camera with pitch tilt and zoom can be used but this reduces the detection capability and detection coverage. Vision systems when set to be sensitive, lose specificity so more false positive alerts are generated. These false positives require operator discrimination to sort detections as drones or as false positives and ensure only the drone events are alerts and logged. If high specificity is set, then the sensitivity is reduced, this can mean false negatives where drones present are not detected and will mean a reduction in effective range of detection and possible increases in detection times. In relation to the precision of vision systems if a drone of a type not seen before by these systems is seen, this may not trigger an alert. In addition if only a single camera location is used, information beyond the bearing will not be reliable so information on distance, height and drone type may be incomplete. Vision systems which include thermal and near visible infra-red cameras can be a very effective at confirming detection by other systems.
Radar techniques operate by actively sending out pulses of radiation and observing the returned pulses. These working for drone size detection are also line of sight systems. They again rely on data intensive processing of the data to pick drones from the other returned signals. This is less intensive than with full video signals but again only with high resolution do you get enough information to have good sensitivity. As with vision systems if you force high sensitivity there can be lower specificity and an operator with access to another system will need to verify and identify the detection. The radar systems most widely used have a limited angle of detection above the horizon which is not a problem in some installation sites but can be a challenge if overflying is possible. For precision, the radar system can provide direction and distance but not typically elevation and is able to monitor multiple drones at once. In isolation the radar detection system is very powerful at detection of flying objects of all sorts and typically needs an additional system to verify the detections as drones.
The final technology for discussion is radio frequency (RF) detection. This is widely used with varying levels of system complexity with matching alert complexity. RF detection identifies the communication links between drones and their controllers. These signals actively broadcast by the drone and controller are very distinct and specific identifiers of drones. This identification technique means specificity is not impacted by increases in sensitivity. This allows RF detection as a technique to have both high sensitivity and high specificity. The complex contents of the messages sent between the drone and controller allow the detection to be picked up within high noise backgrounds and very uniquely identified. The RF detection typically works in a full sphere of detection, with varying sensitivity depending on the antennae used. Only large or conductive obstructions will block RF signals and prevent detection so it operates beyond line of site. In relation to precision, RF detection allows for details of the drone and controller to be reported if the drone type is known to the system. For certain detectors, they have access to the detailed flight information such as exact GPS position of the drone, altitude and flight speed. In relation to all the other techniques known, RF also is the only system which offers identification and location information of the drone controller.
Quality Drone Detection
The quality of drone detection with each of the above methods is dependent on many factors unique to each system and for the broadest coverage possible a mix of detection techniques is best. However the operating principle of RF detection allows for multiple drones and their pilots to be detected simultaneously with no need for an operator to verify the alerts through another method. The other techniques all offer fair detection but are both more complex and demanding not only of hardware and the associated maintenance but also in manning the systems with suitable operatives. All of these systems can in theory detect single and multiple drones and alert to these but only RF can detect and locate pilots, in many cases even before the drone is airborne. It is also the only full sphere of detection solution and does not rely only online of sight for detection capability.
All of these above detection methods can be deliberately fooled or circumvented by drones with unusual body types or specific flight paths that can avoid detection by radar and vision systems. Similarly unusual communication types can avoid RF detection. By combining one or more detection technology you can increase the scope of the detection and benefit from the strengths of each of these systems but no system is ever going to achieve complete detection of all possible drones. Knowing this is it best to consider what level of detection matches the level of risk an entity faces.
How to select the investment level and the complexity of any drone detection system to adequately address the risks drones pose is not trivial. Detection systems are evolving and the drone field is changing radically every year. Any assessment made now should be revised periodically to reflect the changing drone environment and consult with industry experts to ensure the information they have is current
To begin an assessment knowing what the activity level of drones are and what risks these pose to your operation would be a good starting point. With this knowledge determining a suitable drone detection and identification system can begin. In terms of a specific solution is a permanent drone detection solution necessary or is it only really of value when undertaking high risk activities? If there are points in your operations where you are particularly vulnerable to drone threats would a temporary solution fit the risks better. For venues with specific events and seasonal requirements a fully monitored permanent solution and the costs of this may not match the risk well. For operations which risk being compromised by specific drone activities are their specific prevention measures which remove the need for drone detection? If permanent drone detection is required, what detection system, using which technology or technologies is best matched to these risks. Does detection alone provide the opportunity to and mitigate the risks posed? Do you need to consider changes to your operations or inclusion of effectors to properly address the risks?
Drones are only going to increase in usage from today forwards and whether used by your government, your business or you personally, they will impact you. Hopefully these impacts will be positive but history tells us there will be risks and hazards associated with any revolutionary technology. Having a suitable plan in place today to use drones well and avoid the risks drones pose to you is a responsible stance.
To find out more about assessing your anti- drone needs you can request a call back from myself, Alex Roskoss or one of our counter-UAS strategists here.