You’ve heard of the cloud – now here’s the edge
The aircraft industry has a problem. Planes are now fitted with dozens of sensors, but in the future they will use thousands. Sensors will track engine performance, fuel usage, wear and tear of parts, and umpteen other things.
It’s an incredibly valuable thing to do. The sensors help identify problems before they become an issue. Engine parts are replaced before they break, at a convenient time and location. Performance can be optimised. Passengers benefit from fewer delays. And the airline’s bottom line gets a boost too.
The tricky bit? A new engine from Pratt & Whitney has 5,000 sensors that generate up to 10 gigabytes of data per second. Therefore a single twin-engine aircraft with an average 12-hour flight time can produce up to 844 terabytes of data. That’s an awful lot to upload to the cloud for processing.
So airlines are thinking about processing the data on the plane. Install local computation and the data can be analysed locally. Connectivity speed ceases to be a worry. And airlines and engine manufacturers can add more sensors, knowing there’s no upload bottleneck to worry about.
The tactic is red hot right now. It’s called processing at the edge. For many readers this will seem like a U-turn. For years we’ve been told that the cloud is the way to go: send data to a remote data centre and enjoy economies of scale and on-demand computing power. In fact, the cloud still enjoys those advantages. But there are powerful arguments for edge computing too.
There’s the bandwidth argument. Any internet of things (IoT) network which records large volumes will struggle to send unfiltered data to the cloud. Oil rigs, factories and industrial sites struggle with transmission. A switch to edge computing solves the problem and latency is improved giving very fast responses.
Driverless cars need instant decision-making powers. Waiting half a second for data to arrive from a cloud service may be too late. Even 5G, which will improve bandwidth and latency, won’t be quick enough and will add to cost. Hence driverless cars will use edge computing. Think of autonomous vehicles as mobile data centres. Any application which needs fast response data services can benefit from edge computing, which will be more efficient than transmitting to a cloud-based facility.
Cloud will continue to play its role, but edge computing is going to be a huge deal. It unleashes the IoT. Computing at the edge enables more sensors to be used and allows you to collect more data. Bandwidth and latency limitations are significantly reduced. And there’s a multiplier effect to consider. Edge and IoT are not used in isolation. They are pieces in a bigger jigsaw. Industries are embracing robotics, artificial intelligence (AI), biometrics, 3D printing, GPS, smartphones, and augmented and virtual reality.
Computing at the edge enables more sensors to be used and allows you to collect more data
A new UK government-sponsored report called the Made Smarter Review expressed it perfectly: “Emerging technology breakthroughs in fields such as AI, robotics and the IoT are significant in their own right. However, it is the convergence of these industrial digital technologies that really turbo-charges their impact.”
Each component feeds off the others. Edge computing can combine sensors with AI. Robotics can be improved with data analytics. The benefits but also the effects are multiplied. The volume of data created and the compute power required both increase massively.
The report, put together in partnership with more than 200 organisations in the UK, including businesses and universities, spells out the end result. The impact of industrial digital technologies on the UK economy over the next decade could be as high as £455 billion and add 175,000 jobs. Industrial productivity can rise by 25 per cent. These are mouthwatering figures.
There will be obstacles to this digital revolution. The Made Smarter Review, which is worth reading in its entirety, lists a dozen. Cybercrime is a threat. Intellectual property is a hot issue. Access to finance and incentives to encourage adoption of new technologies is needed. But perhaps the hottest is the access to skills. Do we have the right people to make the most of industrial digital applications?
To make the most of edge computing and the new revolution in tech we need to be honest as there is a skills gap in the UK. The World Bank estimates the country needed an extra 745,000 digital workers between 2013 and 2017. It estimates 90 per cent of new jobs will require digital skills to some degree.
Training is one solution, including investment in both education and lifelong learning, but companies need to think digitally and act smarter.
HOW TO MAKE IoT SMART
One example of thinking digitally and acting smarter is a new platform, launched by Hewlett Packard Enterprise, that radically simplifies application deployment in manufacturing plants and leverages both edge and cloud technologies.
The Express App Platform – Manufacturing makes it easy to digitalise industrial operations. With just a few clicks, a variety of industrial applications can be installed to streamline operations. The platform is located on-premises. This intelligent edge deployment means there are no latency or bandwidth issues. The platform works with the cloud via HPE’s Cloud28+ catalogue, which lists the applications. It becomes straightforward to adopt groundbreaking digital applications, such as manufacturing execution systems or advanced planning and scheduling, as well as IoT applications.
The headline for manufacturers? I think there are two. The first is to investigate how edge computing can enhance your organisation. It’s new, but it’s a game-changer. The second is to assess the impact of Industry 4.0. It’s beyond simple optimisation. It’s about transformation.
The airline industry is a great example. IoT sensors reinvented the way technicians fix engines. It changed the way engines are sold as most are now rented under a service agreement and the servitisation model. Both are a clear break with the past. Now other industries must find their own step-change implementations.
We have incredible new technologies at our fingertips. And we have easy-to-use platforms for accelerated deployment. The potential gains are huge. The government has shown impressive foresight in making industrial digitalisation a key focus of the new Industrial Strategy, but now we need decisive action by business and government together to take its key recommendations forward.