What were the biggest Industry 4.0 advances in 2018, and what’s next?
6 minutes | 17 Dec 2018
Edge computing and the convergence of big data with other technologies have taken us a step further than where we were in 2017.
Big data, IoT and cloud computing are each their own disciples that have become intertwined with each other. The demand for big data led to adopting both the Internet of Things (IoT) and cloud platforms. With IoT, the amount of big data naturally increased. The next step? With IoT and big data came a move towards cloud technology.
The use of Machine Learning, really a subset of Artificial Intelligence (AI) gained momentum as well. If you shop on amazon, you’re already familiar with machine learning, with online recommendations based on what you’ve browsed and bought in the past. Machine learning simply trains a machine how to learn to perform human tasks. A self-driving car is another example.
The growth of machine learning stemmed from the ever-growing volumes of data. In 2018, more of us used algorithms to produce models that analysed bigger and more complex data. In turn we got faster and more accurate results without having to intervene.
Applications for Digital Twins also spread across manufacturing. While this technology has been around for more than 15 years, IoT has made it more cost effective to implement. Simply put, a digital twin is a complex, virtual model of a physical object. It is that object’s exact counterpart in digital form, in other words. The object could be anything from a car engine to a building.
Sensors are connected to the physical object and sent back to the digital twin. The data collected from the connected sensors telling engineers, in the words of IBM, ‘not only how products are performing, but how they will perform in the future’.
Also known as Fog Computing, Edge Computing refers to the computing infrastructure that exists close to the sources of data, that is, your machines. The data is processed near the edge of your network – where the data is being generated – instead of being sent to a data centre.
There you have it. In edge computing, data is processed by the device itself or by a local computer or server.
What’s the advantage? It’s about speeding up the streaming of data. Think of a self-driving car, which manufacturers took to new heights in 2018. With the data being processed by the car itself, there’s no lag time. The car can respond immediately to what the data is telling it.
On a factory floor, edge computing gives manufacturers the benefit of reduced bandwidth usage. There’s also a measure of security involved, as data doesn’t have to go onto a public crowd.
We’re adding this to the list even though its existence is called into question – in its intended form – though IBM, Google and Intel have built simple quantum computers. Quantum computing received a lot of attention in 2018. It’s a tough nut to crack. You have to start with quantum theory, which is a branch of physics focused on atoms and the smaller particles – subatomic – inside of them.
Atoms don’t behave according to the laws of physics as we know them. Richard Feynman, one of the greatest physicists ever known, once wrote: "Things on a very small scale behave like nothing you have any direct experience about . . . or like anything that you have ever seen."
These tiny atoms represent what’s at the heart of quantum computing. Where bits are a key feature of traditional computers, called ‘classical’ computers, ‘qubits’ are a feature of quantum computers.
In Forbes,Bernard Marr sums it up: “a computer using qubits can store an enormous amount of information and uses less energy doing so than a classical computer. By entering into this quantum area of computing where the traditional laws of
physics no longer apply, we will be able to create processors that are significantly faster (a million or more times) than the ones we use today.”
Why is quantum computing so important? According to a report by the Semiconductor Industry Association, by 2040, we will no longer have the capability to power all of the machines around the world. You can see why governments and the
computer industry are working around the clock to make quantum computers work. Quantum computers are not intended to replace our laptops or desktops but will be used on a commercial scale.
IBM has a good headstart. They’re even imagining the possibilities for quantum computing’s applications:
- Supply chain & logistics: efficiencies can be found that we can’t yet imagine
- Artificial intelligence: making machine learning much more powerful
- Medicine and materials: can lead to new discoveries by simplifying complex molecular and chemical interactions
- Financial services: the ability to isolate risk factors to make smarter investments
What will we see in 2019?
Mostly, more of the same of what we already have, but better:
Chatbots will get smarter
If you’ve ever been frustrated by a chatbot, know that it will get better. Natural language processing (NLP) will improve thanks to machine learning algorithms. If you’ve not heard of NLP, it’s what allows technology such as Siri or Alexa to understand what you’re saying and react to it. In 2019, NLP will get to a stage where you’ll find it difficult to tell if you’re talking to a flesh-and-blood human or a computer.
NPL will fundamentally change how businesses provide services, from fast food ordering to banking. What about bad grammar in speech or bad spelling when talking to a chatbot online about a product you’re interested in?
Advanced NLP will be able to understand what the speaker or writer means – it will understand your intent. Right now, if it’s not clear if you’re asking a question or making a statement, you’ll only confuse the chatbot. That will soon be a problem of the past.
Robots will become more sophisticated, becoming smaller and smarter, says a report by CB Insights.
Expect to see more robots in use in advanced economies. They’ll also become flexible, learning new tasks as needed. This is in large part thanks to AI, which is making robots easier to train (which makes them a feasible investment for smaller companies with strict ROI needs).
Right now, some robots can be taught a task by guiding their arms. It only takes a few times to demonstrate to the robot what to do before it can program the motion itself. It’s a perfect fit for companies who need that kind of flexibility.
Access to 5G
5G networks will be able to handle more data and connect more devices all at once with much faster speeds than allowed by existing technology. It’s already available in the U.S., though not everywhere, but many telecoms are rolling it out in 2019. The UK won’t be far behind if all goes to plan.
Only 77.28% of network users have access to 4G. BT– British Telecommunications – wants to roll out 5G to heavy 4G users first, aiming for the end of 2019.
The rollout of 5G isn’t just great news for mobile phone users, but for manufacturers and other businesses looking to ramp up or adopt IoT. With data transmission times up to 10 times faster than 4G, 5G will make real-time data more reliable. Device-to-device communication will be faster than we can imagine. Think of self-driving cars and how 5G would apply – faster data feeds will provide an important safety factor by preventing collisions. The car will ‘think’ and react faster.
New infrastructure will be needed, because. 5G operates at a frequency that proves much higher bandwidth than 4G requires. Because investment in that infrastructure is part of the story, 5G won’t be a fact of life for everyone, not in 2019, anyway.
Blockchain will mature
Ajit Prabhu, innovation Leader at Deloitte, doesn’t foresee a widespread adoption of blockchain in 2019, but rather, interest in identifying use cases to understand its scalability and performance. ‘There is increasing interest in corporations understanding what types of business problems are best suited to blockchain,’ Prabhu notes. This will involve learning by trial and error, proof-of-concepts, and other experiments.
Lucidity’s COO Nikao Yang considers scalability ‘the number-one hurdle blockchain has to overcome in order for developers to build apps that solve real-world business challenges.’ Yang also points out that while it’s early days, sidechains – a separate blockchain attached to its parent blockchain – appears to be addressing the scalability issue.
Mark Smith, CEO of blockchain development company Symbiont, told CNBC that after four years of research, they’ve found that the technology works best when there's an opportunity to automate processes, such as data collection.
"I think supply chain is going to be the first, if not near the first, to really show the value of blockchain," Smith said. ‘There aren't any regulatory questions in supply chain management that you have to deal with.’
Trust is blockchain's real value, according to supporters, who, according to CNBC, ‘say the technology will work best not as a new system within private enterprise, but in its pure form — an open-source, publicly accessible platform powered by a global network of users.
According to Altman Vilandrie; Cowen and Company, we won’t see widespread adoption of blockchain for another 5.9 years.