January 30, 2018. Technical Edge and IoT Author: David Fearne

What are the Hottest Technology Trends for 2018 - Part 2

If you missed Part 1 of this post make sure to check it out.


AI Hacker


5. Offensive AI

Sadly, one of my security predictions for 2018 is not so great. One of the key trends we have seen in 2017 has been a dramatic rise in large scale and state sponsored cyber attacks.

As of today, all of these attacks (although clever) have leveraged an existing vulnerability in software, hardware or processes.

This has led to patches or process changes and then we continue, albeit with our tails between our legs.

Make no mistake though, the business of hacking, writing malware or generally exploiting systems is as large as the business of protecting them.

And they are also using the same advanced technologies as “the good guys” to create ever more intelligent attacks.

Enter the possibility of ‘Offensive AI’. Now I’m not talking about Skynet here but what I am talking about is the ability for attacks to become super intelligent, to adapt themselves, to hide, evolve and learn as they move around a system.

Imagine training an AI to think like a hacker while inside a compromised system? It really could be the next major battle ground for cyber security.


Quantum Computing Abstract Concept


6. Quantum Computing Goes Mainstream

We need quantum computing! Why? I hear you say…

Because if we don’t figure out another way to compute data that’s faster than we have today, progress will stop.

Why? I once again hear you say.

It’s because we have done such a good job of keeping on track with Moore’s Law.

This is the observation that the number of transistors in densely integrated circuits will double approximately every two years via advances in the miniaturisation of the transistors (the bits that make computers work).

However, the laws of physics wont let us go any further.

The more transistors, the faster the speed of the processor, the smaller the transistor the more we can fit, meaning the faster we go.

Simple, until we get so small that the transistor isn’t big enough to actually stop the electron it’s trying to route in a certain direction and that electron just breezes past (quantum tunnels) the transistor.

At this point the processor fails and the computer wont start anymore. So the old phase if you cant beat them join them comes into play.

Physicists have long theorised the possibility of using this odd property in computing and in 2017 it all got a lot more serious.

IBM started a division called Q-Series and released a SaaS interface to their own quantum computer, Microsoft released a quantum development kit, and D-Wave sold a 2000 Q-Bit system to a cyber security firm.

2018 should see commercially available software applications that use quantum computing as an accelerator.

As quantum computing is not currently a general-purpose computational platform, this will be to provide quick and more accurate results to optimisation problems such as scheduling and routing where millions of combinations of options need to be taken into account to get an answer.

This is currently a very slow and compute intensive process, with quantum computing every route, option or condition could be processed at once, instantaneously.




7. Harder, Better, Faster... you need automation

The more we rely on technology to do everything in our business, the more we need to leverage the connectivity of machine to machine via intelligent processes.

That means more decisions, functions and tasks can be carried out repeatedly, accurately at a lower cost and a lower time frame.

Ok, prediction done… Or is it?

Sadly not, the above statement sounds so ridiculous that I’m sure you’re thinking ‘well, why isn’t everyone doing this?

Simple, because automation, and I’m referring to automation of many systems in unison, as is inevitable whilst delivering business function in a real world enterprise, is hard.

Not because the tooling isn’t available or matured, but because the interfaces or APIs that softwares expose for automating have no common format and change regularly with updates of the system.

And, much like when you get a human to navigate a new interface of the software they use on a daily basis, this task will break the automation and need professional intervention to repair.

This has put many users off as it can sometimes seem like a juggling act when once you have one systems integration up to date, another one changes its interface.

This can be minimised with strong process orientated change control, but in 2018 I predict that we will see software vendors start to put more development into integration modules; these are the glues that stick them to a business process automation system.

This means that when new interfaces are developed, the overhead on the organisation will be minimised, if not completely removed, and they can, as was initially intended, focus solely on building the best set of automated business rules to help them work harder, better, faster.




8. Manufacturing the Future

Industry 4.0 is a name for the current trend of automation and data exchange in manufacturing technologies and is broadly defined as; cyber-physical systems, the Internet of Things, cloud computing, and cognitive computing.

Industry 4.0 creates what has been called a "smart factory".

The smart factory has the ability to optimise its production line, production characteristics or level of human involvement autonomously, drawing on an internet of services model such as cognitive computing, automation and cyber-physical systems such as robots and connected 3D printing systems.

Another element of Industry 4.0 is a concept called digital twins. This is a digital representation of a physical object, from the material level up, to allow for product prototyping and innovation to occur at a much faster pace.

It combines artificial intelligence, machine learning and software analytics to allow the twin to act in a virtual environment exactly as if it were a real physical copy.

This dramatically lowers the cost and time required to create and iterate products, as well as lowering the possibility of faults due to more rigorous, automated testing procedures for every iteration of a product during the design phase and into its use lifecycle.

So why is this a trend for the channel unless you are in the manufacturing industry vertical?

Well, all of the above requires a very highly skilled workforce rather than the mass workforces that global manufacturing currently uses.

To that end, many could envisage new cyber factories being built in Western Europe and the UK due to the high availability of these skills.

This would, in turn, open huge opportunities in infrastructure to house these incredibly compute intensive factories.

Not to mention the opportunities for supporting and implementing the combination of solutions that will need to go into the factories themselves.

So, get yourselves along to an Industry 4.0 summit and contribute to the next industrial revolution becoming a reality. 




9. LPWAN technologies will open up the IoT market

We started with Local Area Networks (LAN) then we had Wide Area Networks, followed by WLAN, SAN, MAN, PAN, CAN, SDWAN (one of last years predictions) and now we have LPWAN.

LPWAN stands for Low Power Wide Area Network. This is directly referring to the wireless technology used in industrial IoT solutions.

IoT doesn’t use WiFi as it’s very power hungry and when you are running off a battery or solar panel power is everything.

So we have created wireless technologies like Sigfox and LoRa to communicate with these field based technologies.

Another limitation of WiFi is the distance it can be used at. In a built up area it can be down to tens of meters.

Sigfox and LoRa use very low frequency radios to communicate with IoT devices which allows for communication distances of tens of kilometers from a single base station, not to mention the ability to penetrate deep into buildings.

There is a trade off - low power means low data rate. The average communication rate of a 6MB per day.

This means that you need to be frugal with your transmissions, but when you are just a temperature sensor this is fine.

The major current limitation of this technology is the availability of it. LoRa is normally self-deployed as part of a specific IoT implementation and Sigfox is a telco lead service but had a false start previously.

It’s now in the UK, being lead by WND who are aiming for 95% UK coverage with around 1800 base stations deployed by 2018.

There is also a new challenger in the market called Narrow Band IoT or NBIoT. This will be a service delivered by telcos to try and accelerate the adoption of enterprise IoT.


There are many challenges and massive opportunities for the enterprise and channel. I’m looking forward to seeing how well businesses embrace these and what unfolds in 2018.

Related Posts

  • What are the Hottest Technology Trends for 2018 - Part 1

    David Fearne & his Technology Trend Predictions are back! Always a popular post to start the year, David shares what he thinks will be the hottest trends in the tech world in 2018.

    > More

  • Arrow Bandwidth | 2017 Tech Trends in Review - Part 1

    In this episode of Bandwidth David and Rich review David's 2017 technology predictions to see how accurate he was in predicting the top trends of the year. Part 2 is up next week.

    > More

  • What are the Hottest Technology Trends of 2017? Part 1

    David Fearne, Arrow's UK&I Technical Director, is back with his technology predictions for 2017. Find out what he thinks will be the hottest technology trends of the year.

    > More

Blog Updates

Popular Posts

Challenges and Solutions for Managed Services Providers in IoT

> More

VMC NetApp Storage

> More

Bandwidth On Point Takeover | Episode 1 - IoT Foundations

> More

Being a Disruptor in the Market

> More

Dell Technologies World - "Make it Real"

> More

Blog Archive