Features

Virtual Borders – Are they really a thing!

by Tony Kingham, Editor, Border Security Report

Whilst the ‘wall or no wall’ border argument continues, developments in smart border technology grows a pace, primarily driven by the potential of Artificial Intelligence (AI) and Machine Learning.

Governments and border agencies across the world, but particularly in the US and Europe have been quick to recognize the potential of AI to dramatically increase the effectiveness of their existing and future surveillance systems.

The so-called ‘smart wall’ linking ground systems such as radar, seismic sensors, fibre optics, cameras, aerostats and drones, with communications and databases promises much, but where are we in terms of delivery?

The main problem for many border agencies worldwide, is that they already operate a wide variety of legacy surveillance systems which may have their own operating systems that do not necessarily talk to other systems in their surveillance environment.

So, it is entirely possible, if not usual, to have a situation in which one team operates the ground-based radar, another team, the electro-optics, others the airborne assets (if they are lucky enough to have them) and yet more on ground sensors like seismic sensors and microwave fences. This creates a problem of cooperation and co-ordination (that may be familiar to many of our readers.) Each team may even be on separate communications systems, which is especially true of airborne assets.

The trouble is, that these legacy systems represent a considerable investment and governments are reluctant, especially in the current financial environment, to bin perfectly good systems and start afresh.

In the past, major projects like a nations border were the domain of the big integrators, many of them from the defence sector. These manufacturers would create a bespoke command and control system which necessarily gave them a lot of influence over the project and what systems could or couldn’t be included.
The promise of open architecture and AI is that it puts the customer back in charge and allows them to choose the what, the when and where, of systems and equipment they include.

So, what is open architecture and why is it important?

Open architecture is software that is in the public domain rather than proprietary closed architecture developed by a single developer or manufacturer. Closed architecture is a closely guarded secret. It does not allow anyone outside of that manufacturer to develop or modify it in any way. The classic example of this is “Apple”.

The exemplifier of open architecture is of course the Android Open System Platform (AOSP) a publicly available and modifiable Android source code. Anyone can download it and modify it and develop new phones and new applications that are fully functional on all other Android devices.

This revolutionised the mobile phone and devices market. It meant that instead of customers relying on a few manufacturers to bring new products and applications to the market, the lower cost of development meant more products coming to market at lower prices.

Just as important the process of software and hardware development was, if you like democratised. Anyone can develop applications for use on any Android device. This dramatically increases the capability of these devices to do everything from sending encrypted messages, to identifying images, constellations, birds or songs, or even measure your heart rate.

The hope is that an open architecture approach will have a similarly revolutionary effect in the area of border security. Allowing border agencies to build capability incrementally, utilizing existing systems and equipment as well as seamlessly adding new systems of their choice, such as counter drone systems, surveillance drones, aerostats and even ground observation satellites.

The goal is to create one operating environment in which all assets are managed by a single command and control system that uses AI to identify a target approaching to the border, decide what that target is, whether it is a vehicle, drone, human or just an animal, and autonomously make the appropriate response. Whether that is sending a human patrol to intercept or sending an autonomous drone to confirm identity or as a deterrent.

It may sound like science fiction, but it is very much a current reality.

AI can handle multiple threats, it never sleeps, and it learns on the job!

Fundamental to this approach is the communications system along with all the necessary infrastructure that it requires, without which it is simply not possible, especially on extended borders like the US/Mexico border or Europe’s external borders. But the infrastructure for new systems is expensive, requiring hundreds if not thousands of miles of power and communications lines, substations, maintenance etc.

To solve this particular problem the U.S. Customs and Border Protection (US CBP) Agency has gone for the obvious and most cost-effective solution in that sun baked environment, which is solar.

Having reliable sunshine makes use of self-sustained surveillance towers powered by solar, a practical and economic way of increasing security on the southern border.

So, into this domain stepped US tech company Anduril, which developed its Lattice open architecture operating system. The system is designed to integrate sensors from third party vendors into a single digital environment where AI/ML techniques can be applied to ease the burden on multiple operators. But a software-only approach didn’t make sense to address all of the issues in austere environments. The lack of civil infrastructure, poor connectivity, and a general topology that prevented border forces from maximizing the benefits of traditional sensors.

Anduril’s software-first, hardware-enabled approach addressed this by using low-cost third-party sensors that delivered enough data to be passed through and retrieved for users to interrogate.

Consequently, they developed the Sentry Tower, which, like products on the consumer market – has undergone several generations of development, continuously improving its capabilities. As a robotic observation post, it uses affordable radar, cameras, and sensors, combining and analysing the data locally. It is solar powered and has more local computer power than most military systems. CBP combines them via radio and other communication carriers to create a hive mind of robots that cover its southern border’s undulating terrain. Anduril subsequently was deployed to other borders and sites to ingest legacy surveillance systems and develop a mission engine that provides border officials with an effective and efficient means of orchestrating resources by serving as a panopticon.

Paul Hollingshead, Head of Europe & Africa at Anduril said, “The objective is to generate continuous improvement, increasing value or reducing costs per mission. Most of our products can be delivered “as a service”, which reflects subscription services in a commercial setting. The government has no sunk costs in developing the system but subscribes to its capabilities. It is delivered via hardware, much like the routers you get from a broadband service provider. The value is in the service, not in the hardware. Customers will benefit from new features and capabilities in both software and hardware during the subscription period. As a result of subscribing, the government encourages the supplier to continually develop or risk losing the contract.”

Anduril have already successfully integrated other OEM systems such as the Echodyne radar widely deployed on the US border, and electro optics like those of FLIR.

Other systems that can be integrated are fibre optics like the Sintela ONYX™, Linear Ground Detection System currently deployed on the US border, as well as C-UAS systems like Anduril’s own cUAS system and Blighter’s AUDS.

In Europe, Monaco based MARSS Group produce the NiDAR C2 system, which again uses an open architecture and is ‘sensor agnostic’ meaning it can integrate existing and new equipment. MARSS have won a number of significant contracts to protect critical infrastructure, especially in Saudi Arabia, but I was unable at this point to confirm whether it is currently deployed on any border.

This sort of approach means that not only can border agencies integrate their legacy systems but they are also able plug-in new systems as and when budgets allow. This gives them the opportunity to use best in class systems where necessary or budget systems where the tactical circumstances allow, thus saving money.

The use of AI allows agencies to better utilise their most valuable and most expensive assets to best advantage, i.e. their personnel! Smart systems will allow personnel to be transferred from operations rooms to field operations. And, once in the field, instead of spending hours in often fruitless routine patrolling and manning observation posts, they can be accurately directed to border interdictions, making them more efficient and saving that most valuable of resources, time.

In a world where illegal migration is a problem that is not going to be solved any time soon, smart borders offer the best way forward.

But of course, a good high sturdy fence may also still have its place, after all, layered security is always best!