The fourth industrial revolution is now upon us, combining traditional manufacturing and industrial practices with digital technologies including advanced computing solutions, cloud and remote computing, Internet of Things (IoT) and connected devices, and artificial intelligence (AI) and machine learning.
Just like the previous three industrial revolutions – steam and water power, electricity and assembly lines, and computerisation – it will transform how things are made.
Speaking at a recent conference on new industrial technologies, Siemens’ head of automation, Simon Keogh, said that edge computing and AI in particular will play an increasingly important role in manufacturing.
Edge computing is when data is analysed on the edge of the network where it is generated, rather than in the cloud.
One example is an autonomous vehicle, Keogh told B2B media company Insider’s Northern Powerhouse Advanced Manufacturing Conference in Greater Manchester: “Consider this as a machine with an on-board edge computer: I would want to make sure that while I was in that vehicle that it’s the on-board computer that is controlling the vehicle and making the decisions.
“I wouldn’t be that comfortable if it was all being managed in the cloud and it was the cloud deciding whether I am turning left or right or making an emergency stop because somebody has stepped out in front of the vehicle.”
In a factory, edge is the layer in between the machines (and their sensors and controllers) and the cloud.
“We’ve realised that the sheer volume of data that’s being created at the control level is massive,” Keogh said. “It simply isn’t efficient to put all that data into the cloud. There is a need for a localised, high-speed, pre-processing layer that sits between the control level and the cloud. Edge technology fills this gap.”
Keogh also explained how advances in AI are enabling manufacturers to automate a wider range of processes, not just those that are predictable.
“Traditionally,” he explained, “you’d automate those processes that are repeated time and time again. Where AI comes into its own is the ability to automate the unpredictable. We’ve developed a neuro-processing unit that sits inside our standard automation controllers. We can use it to train robots rather than programming them.”
This allows for greater flexibility: we’ll tell the robots of tomorrow what to do, and they will decide how they do it.
A new report from Osborne Clarke explores the impact of IoT within the manufacturing sector in Asia. Download your free copy of How will IoT transform industry in Asia? to learn more about how the manufacturing and logistics sectors in China, Hong Kong, India and Singapore are embracing IIoT technologies.