Artificial intelligence is the beginning of a revolution, but in one way it is just like every other revolution: It can be abused. Whether or not you already use any AI, you need to understand two things; that AI is cranking up the severity of security threats, but it can also offer improved security.
AI systems are fast and dynamic, meaning they learn from experience instead of relying on pre-programmed assumptions. AI-powered malware doesn’t require the hacker to know anything about you in advance. However, an AI-powered defence system needn’t depend on fixed definitions of who to trust and who not, or how they gain access. It can learn to recognise suspicious activity.
AI will power more advanced intrusion attempts into systems that are themselves more powerful. End users need to understand that the sophistication of AI-powered tools does not mean they are secure. For example, facial recognition systems powered by AI can potentially be spoofed by another AI, providing building access to criminals, or framing innocent people with forged video footage.
A report from Forrester “Using AI for Evil” says “mainstream AI-powered hacking is just a matter of time” and Ciaran Martin, of the National Cyber Security Centre, said it’s a matter of “when not if” [there will be a major attack on the UK].
Using “bot manipulation”, malware can use AI to adapt its appearance so that antivirus software doesn’t recognise it. It can also use AI to sample normal network activity and use it as camouflage, known as “generative networks”. When the target is itself an AI system, a malicious actor can feed “poisoned” data to the engine in order to bypass filters or simply cause damage. AI can also learn to impersonate a legitimate person or company in order to launch a social engineering attack.
The ability of AI to react quickly and adjust its responses as situations evolve also makes it ideal for defenders. An AI security system gives defenders the edge by providing early warnings and rapid incident response, so attack vectors can be closed down before any real harm can be done. Darktrace is one such tool.
Behaviour analytics is another important defence tool. Detecting unusual activity allows the AI to close access to key resources while a deeper examination is undertaken, for example, using Varonis. Mastercard’s director of cyber and intelligence solutions in South Africa, says AI is saving $20 billion per annum by detecting fraud in this way. Embedded malware code can be detected using a similar method.
AI-powered solutions also help by improving activity logging; centralising it in a single place and providing tools to zoom in on significant trails. The logs collected by Azure and other Cloud platforms provide a good basis for an effective SIEM system. These tools also enable you to create and evaluate your alert response workflows.
Once in the Cloud you have access to specialist security products and expertise that few enterprises can deliver in-house. Specialist companies constantly monitor the global situation to stay aware of threats emerging in particular sectors or locations. An ideal SIEM integrates this digital intelligence with your standard procedures such as logs, asset inventory, AI pattern detection and automated incident responses, and makes it easy to demonstrate your statutory compliance.
Unfortunately, we can’t wait for someone else to solve our cybercrime problems. The very people we should be able to trust to protect us, the NSA and GCHQ, created the EternalBlue tool used in recent ransomware attacks such as WannaCry, NotPetya and BadRabbit. They also left exploitable flaws in Windows and implanted backdoors into server and router firmware. Although this is similar to the warnings against Huawei, the NSA have placed similar backdoor access into products from Cisco, Juniper and Fortinet.
The problem with creating these weapons is that everyone else soon uses them; innocent companies are the victims. According to Wikileaks on 7th March, the CIA regularly listens in on Samsung televisions and iPhones and can take control of numerous IoT devices and car computers. When they do it, others will soon follow.
For businesses the goal is clear, keep spyware and vulnerabilities out of your software and hardware. That means taking a keen interest in where your IT products come from and investing in good security. There are limits to what is practical, but an integrated security system powered by AI is the best possible solution
5G has suffered bad press from both detractors and supporters. Spoof stories about it spreading coronavirus were soon dismissed, but banal predictions of refrigerators ordering milk and shoppers wearing headsets to receive advertising were even more likely to blunt our interest. 5G undoubtedly creates the groundwork for an enormous technical revolution but adjusting the central heating with our smartphone or watching B-movies in higher resolution is not the point. Manufacturing and logistics industries will lead the real 5G revolution.
Although the public 5G network will take some time to get up to speed, local area networks can implement true 5G more quickly. This will enable factories, ports, universities, farms and airports to have their own industrial IoT systems (IIOT) today. Numerous factories are already claiming the ‘first’ 5G production lines, including a Nokia factory in Oulu Finland, Worcester Bosch in the UK, Mercedes Benz in Sindelfingen Germany and General Motors in Michigan.
Speed is often mentioned as a key advantage of 5G, but it helps if we break down the meaning of ‘speed’. 5G radio waves don’t move more quickly than 4G ones, rather the entire system has been optimised for faster data transfer. 5G can reduce latency to as little as a millisecond, enabling machinery to respond to sensors almost instantly.
Consider how quickly a driverless car must respond in order to operate safely and you will understand the value of low latency. In a similar way, 5G will enable a whole new generation of robots and automated machinery to radically improve dexterity, quality control and safety. Ericsson’s vice-president Åsa Tamsons explains:
"With one millisecond latency, you can sense whether there is a deviation in the process before the tool even hits the blade and you can stop the machine before the error happens".
‘Edge’ responses in today’s driverless cars are achieved by mounting the control device directly on the vehicle. 5G cars will achieve similar response times but with all the benefits of environmental network connectivity too.
5G also has far broader channels so that more devices can be connected simultaneously. It is said that 5G will soon be able to connect a million devices per square kilometre. Imagine what an engineer could do with ten thousand eyes and ten thousand hands. All the extra data feeding into AI enabled machinery would provide a precise real-time grasp of complex distributed systems and emergent situations with many industrial applications.
Not all 5G systems need to be this fast, but a typical industrial 5G LAN will match a good Ethernet one. A huge disadvantage of Ethernet is the wires, they are expensive to install, prone to breakages and need regular maintenance. In contrast, once setup a wireless 5G system is easy to maintain and reliable (99.9999% or ‘six nines’ reliability).
One reason for hard-wiring a system rather than using ‘wi-fi’ is because most types of wireless connection can fail to penetrate walls and metal obstructions. However, 5G is relayed between multiple small nodes and can re-route itself instantly if a passing tanker or crane blocks any particular path between devices. The technology is called ‘coordinated multi-point’ (CoMP).
Finally, 5G provides much improved network control, including the ability to subdivide the network. Known as ‘network slicing’, this means each virtual sub-net can be customised and optimised for multiple different purposes.
Whether public or private, 5G networks have applications everywhere. By planting sensors in the ground, farmers will know precisely how much water or fertiliser their crops need and when, or query weather satellites and predict their ideal harvest time and yield. Driverless machinery will often deliver it. The health of herds can be monitored remotely and assets tracked across the farm and supply chains.
The IoT has already demonstrated multiple applications in health and fitness. We are beginning to use proximity sensors and temperature sensitive cameras to track disease outbreaks. In the future 5G may be able to stop a public health threat in its tracks. Augmented reality may also facilitate remote examinations, benefitting people in isolation and the NHS system.
5G supports three rather different kinds of technology; smartphone broadband, large-scale IoT and critical ‘edge’ operations. Because smartphone makers need to sell handsets to pay for the public network, some of the more frivolous ‘benefits’ have been hyped. Many people will receive a Samsung S20 this Christmas and wonder what to do with it. However, the real revolution will be quieter and more impressive: few enterprises will be able to ignore 5G and still remain competitive.
Never before has it been possible to collect so much data. However, the data is worthless until you can mine it for information, which in turn is useless unless you can understand it. It’s disappointing that most companies are still reliant on two-dimensional charts, graphs and tables of impenetrable figures. The underlying data is labour intensive to collect and processing it can take so long that actionable reports are often out of date.Finally, if you decide to respond in a particular way, you will have a further wait before you can evaluate the outcomes. Visual analytics combines the tools needed to perform all these steps but on a much faster timescale.
Briefly, your information resources can be collected automatically by sensors and cameras or by querying a wide variety of company data resources. Once you have a single point of access, data mining or similar pre-programmed algorithms can rapidly extract and organise it. You will then be able to extract meaningful correlations and aggregate key statistics. At the monitoring end, the salient information is provided in visual forms that human beings can understand at a glance and then respond swiftly.
Although they are complex, visual analytic systems are extremely flexible. If you can gather digital data on the activities or operations you need to monitor, you can apply visual analytic tools to them. This means it has a role to play in business, security, governance and on industrial production lines.
With many tools now available in the Cloud, it is within the reach of small to medium sized businesses for the first time. Many visual analytic systems are configurable using simple drag-and-drop interfaces, so although you need to understand your own operational requirements to design them, you don’t need skilled specialist IT teams to operate them for you.
If you can collect your data in real-time, such as from remote sensors and cameras linked across the IoT, then you can not only respond in real-time, but view the consequences of that response in real-time too. Many analytic suites also enable you to explore the consequences of a policy or production line change before you commit to it, as well as to identify historical trends.
Data science and statistical analysis isn’t something that everyone has time to understand, but pointing and clicking with a mouse is now commonplace. Visual analytic interfaces are designed for operational managers, to help them focus on their own areas of expertise and their own specific issues. Business managers and operational technicians can collaborate to devise solutions without needing to refer to IT specialists or external consultants.
Development is also simple. Sophisticated analytic systems can be built up without ever having to call in a coding team. Your solutions can be built in the Cloud, inside your intranet, or close to critical points in your operational infrastructure. The absence of a steep learning curve means there is a rapid return on investment for the company.
The flexibility of these systems enables, rather than replaces, human insight and experience. There are many areas where you can use these tools, guided only by your own creativity and imagination. In the course of exploring different data views you are very likely to discover answers to questions that you might never have thought to ask.
Before data analytics, if a report made you aware of a problem but didn’t explain the cause you would have to request more detailed information from front line departments and then wait for a further report. In contrast, visual analytics lets you explore a succession of views until you find the one that answers your question. You can dig deeper, or change the way you examine it with a single click until you find the view that makes the answer clear. As a result, your analyses are more thorough, more penetrating and, critically, up to date.
Analytic resources can be used to create a leaner, more agile enterprise, by making your front-line teams and managers more self-sufficient. Visual systems can reveal exactly where your production line bottlenecks are happening, or to predict where they are going to occur in the future. You can then make prompt adjustments to keep production flowing.
Access to a visual analytics dashboard can empower every member of your organisation by revealing exactly how their process is performing and whether it is keeping pace with other dependent processes. It can also be used to track management objectives, such as KPIs and other project landmarks.
The last decade has seen huge advances in artificial intelligence, smart devices and video analytics. The next will see a dramatic increase in the devices built from them. In fact, demand will be so high that we need to start thinking about our capacity to deliver them.
One bottleneck is the networks over which we expect them to connect. As 5G rolls out, 4G is still patchy outside urban areas and the capacity of our networks to carry 5G traffic has been questioned. Its rollout was also somewhat muted by attacks on phone masts by protesters.
Data centres are also feeling the strain. As more companies, individuals and devices link to Cloud services, data centres have to increase capacity, but noise abatement and heat dissipation make expanding or finding new sites a challenge.
The irony is that only a few emerging technologies need an explosively growing network; demand seems to be driven by people rather than machines. Follow any link to a 4G or 5G website and you quickly discover the benefit of being able to download a 2hr movie in 10 seconds. A strange boast considering that almost everyone now streams, not downloads, movies (and we can’t help wondering why they need them on the move).
By comparison, a smart meter reports your gas and electricity usage about six times per day, taking about 3 seconds in total. Smart meters also use data maintaining their network but that only raises their usage to about 1 minute.
Only a few devices need to transmit more than a few kilobytes of information per hour, nothing comparable to a movie download. Visual feeds from cameras are heavier on bandwidth, but how many hours of CCTV footage of empty buildings do we really need to collect on central servers?
The IoT is a outstanding medium for data gathering and remote control; the Cloud is ideal for data storage and leasing advanced applications, but the most exciting frontier is the development of autonomous systems. When we can store sophisticated algorithms on a chip, smart devices are not only less dependent on human management, but also less dependent on networks. Problems such as communication interruption, bandwidth overload, and response latency begin to disappear.
The obvious example is the self-driving car. Not only are they heavily dependent on advanced image recognition but must perform it at a blistering speed. If they had to depend on a remote server for their analytics, they could never match the response times of human drivers. There are several other reasons for providing self-driving cars with a connection (traffic information for example) but the visual analytics that enable it to drive have to be local.
Video feeds are also a heavy load on human observers. CCTV security systems will be more effective when the equipment itself can identify salient events. In fact, the raison d'être for driverless cars is to improve on the situational awareness and sluggish responses of tired human drivers.
Cloud (or other network) dependence is the weak link in many IoT deployments, impairing its speed and reliability. The alternative is to distribute the processing workload close to the edge of the network - near the device. This is often called “Edge computing”.
Rapid situational awareness can often be achieved by incorporating AI or video recognition algorithms onto the device itself, or supplying them in a specialised processing unit in close proximity. This infrastructure can still work in symbiosis with distant resources and control systems, but the bulk of the processing is shifted as close as possible to where it is immediately needed.
In the next few years, real-time information response capabilities will find a multitude of new niches and transform existing ones. For example, video surveillance has been booming for years (in retailing, transport and security systems), but re-establishing those systems on edge architectures will transform their value by making the intelligence they collect actionable.
Knowing which bus ran you over might be useful in an inquest, but we would rather be warned that the bus is coming. Or consider the difference between scouring a police officer’s bodycam footage to see who fired at them, with a system that can recognise a gun and issue a warning that saves their life.
Ideal solutions will often be hybrid. Many systems can learn to recognise faces locally, for example at ATMs and robotic checkouts, yet they can still liaise with central repositories when needed.
Fully autonomous robots are no longer far-fetched, but in the meantime let Net 4 show you how to future proof your video processing systems.
