Articles
Everyday objects such as alarm clocks could be intelligent and connected
Officials from the German government are heralding the start of the fourth industrial revolution. “Industry 4.0”, they say, will radically transform manufacturing and business models.
Previous industrial revolutions have centred on the introduction of machinery, mass production and automation. This one, they say, will see smart factories where machines organise themselves, supply chains automatically coordinate and products communicate with manufacturing equipment.
Germany is so convinced by the potential of smart factories that it is investing ¤200 million in the development of an Industry 4.0 implementation strategy. Although the term is little known outside the country, some of the building blocks that could make it a reality are starting to make it out of the laboratory and onto the shop floor. For GE, one of these is the industrial internet which enables machines to communicate with each other.
Machines running on the factory floor are hooked up with a series of sensors that collect information about their functionality. A sim card or other networked device connected to each of the machines sends this data to a centralised secure server via the internet which takes in the data, processes it and delivers its recommendations.
The recommendations might help manufacturers to optimise the use of materials or better schedule predictive maintenance. The server might also communicate with other machines in the supply chain, for example. The information is given in real time, helping decisions to be made faster, and often describes aspects of production that have been invisible in the past. This new information could enable businesses to come up with innovative products or services.
Carlos Härtel, managing director of the GE Research Centre Europe, explains: “If one person has access to information from a whole fleet of wind turbines or MR scanners, the chances are that with this massive amount of data it is easier and much more effective to pinpoint the performance issues and make predictions about what point in time the system is going to fail.”
Härtel says that the development of high-powered computers, analytical tools, connectivity and the cloud has created “low hanging fruit” that is ripe for industry to use to its advantage: for example, hooking up the assets in a gas-fired power plant to an industrial internet to analyse how fuel efficiency could bring benefits.
According to GE, even a 1% efficiency improvement in this setting could yield savings of $66 billion in fuel costs.
Until now, achieving this type of efficiency saving would require GE investing billions to create a more efficient gas turbine. Härtel says: “Here comes an opportunity to get a percentage point of efficiency without investing millions or billions in new machines. It’s the same machine, only with a software package on it and connected to the internet. We’d be foolish to overlook the opportunity.”
Machine-to-machine communication also enables control of a certain machine to be split down to a number of distributed control devices that operate separate modules, which can be reconfigured and monitored remotely.
Chris Biddle, director of manufacturing systems at Rolls-Royce, believes that manufacturing cannot move forward without the further invasion of information technology. He says: “Tolerances are becoming tighter and we are pushing our products to the edge of physics so we simply can’t rely on human beings to translate that very tight design intent into manufacturing output any more.”
He adds that it is not feasible for Rolls-Royce to continue buying big boxes of metal and machine it away to create a part. “It has to get to the point where we are more or less consuming all of the material.” And this, he says, is only possible with computing capability.
As manufacturing becomes more dependent on digital processes and systems become integrated, the risk that something can go awry increases. But Biddle says the value of having integrated systems far outweighs the risks.
He suggests the issue can be dealt with by adding redundancy into systems, or putting strong business continuity measures in place such as first- and second-line support mechanisms.
Given all the purported benefits of connecting machines to an industrial internet and digitising manufacturing, can Industry 4.0 live up to the hype? Professor Robert Harrison is sceptical.
Harrison is professor of automation systems at the Warwick Manufacturing Group at Warwick University. He says that smart factories have the potential to greatly improve the efficiency of working. But adds: “It’s going to change the way we work but whether it is a new industrial revolution now, I’m not sure.”
Companies that Harrison has encountered are “generally positive” about getting involved with the technology. But some have reservations about its maturity, security and safety.
There can also be resistance to the idea of building systems with virtual tools. But this is changing as more people are exposed to embedded devices and internet connectivity with phones and PDAs, he explains.
Biddle sees similar issues at Rolls-Royce. “There is a greater awareness on the shop floor than there ever has been but there are still people out there who don’t use a computer on a regular basis, so it’s a cultural acceptance issue that we have to get over,” he says.
Manufacturers also have to grapple with the fact that software vendors can oversell capability. Biddle says companies can only find out the true functionality, performance and vulnerability once they have committed to a package. It can then take time to work with the software vendor to mitigate any issues.
As factories churn out data and products, the amount of data available to businesses can be overwhelming. With market estimates pegging the number of devices connected by 2020 in the billions, any single enterprise could be dealing with thousands, tens of thousands or even hundreds of thousands of connected assets.
Jari Salminen, global business development lead at Vodafone Group, says: “The data volumes might become huge.” Getting the data is just the start. Managing it and knowing what to do with it is “critical”, he says. Processing the data effectively relies on algorithms that find structure, signatures and anomalies within it, and using programs that can detect correlations in large data sets.
Vodafone estimates that there are 100 million assets and devices connected globally and, according to Salminen, much of this activity is in the manufacturing sector, centred around remote control and monitoring of assets.
The benefits of machine-to-machine communication are not limited to the factory floor. Proposed EU regulations could see all new vehicles fitted with a built-in sim card that could automatically dial for help in the event of an accident by 2015, for example.
“We are already seeing many different types of asset being smart and connected: shipping containers, individual pallets, individual devices. Parts of a car could become intelligent and tell you where they are,” he says.
In the longer term, it is conceivable that everyday objects such as alarm clocks and coffee machines would be intelligent and connected.
The concept of automated and networked consumer goods is more widely known as the internet of things. For Salminen, machine-to-machine communication represents the first step towards that vision.
In many cases the technology used to connect up machines has been around for decades but, because of its cost and complexity, it has remained in small-scale and niche applications.
Several developments in recent years began to push the technology into the mainstream. Mobile cellular networks, which allow people to connect their machines using existing technology, are now ubiquitous. In addition, the cost of data and hardware has fallen.
Academics like Harrison have been working on web services and embedded control services in a manufacturing setting for several years, and on distribution control and component automation for longer, but so far this hasn’t filtered into the mainstream.
Harrison says the tide will change once big companies fit the technology as standard in their devices. “They are to some extent more expensive but I imagine they will become comparable in cost in the short term,” he says.
Research on uses for machine-to-machine communication is continuing at Warwick. Harrison and his colleagues are working with Ford and Jaguar to build full-scale demonstrators to test whether the technology can give the level of automation manufacturers require.
The laboratory has a 12-station industrial machine, networked using ethernet connections, and researchers are applying remote monitoring and support tools to it. The five-year programme has just begun and will look at the design of automated machines, components and remote monitoring. Researchers aim to quantify the benefits that these new systems could bring compared with current methods.
Among the manufacturers that could benefit most from the technology are carmakers, says Harrison. In their sector, where regulations are regularly revised, firms must frequently reconfigure production systems to keep up. By connecting machines, the supply chain can be kept up to date.
Joachim Klink, director and global automotive and aerospace industry architect at Hewlett-Packard, says that in recent years flexibility has become critical to the automotive sector.
“In each of the last five years, globally we have had a growth or shrinkage of at least 10% capacity, some years 20%. That gives you an idea of how much they needed to add flexibility,” he explains.
That kind of flexibility can only be achieved by having a virtual environment so that once you start production all the parts are in place in the right order, quality and quantities, he adds.
Carmakers were able to do this by using manufacturing execution software that couples machines in their plants with those of the external suppliers, as well as modelling demand, capacities, dependencies and inventories.
Looking ahead, the future could see smart chips embedded in products that are in service that send data back to base about usage. Internet connectivity would enable engineers to gather data on these products to improve operations in the field or during maintenance.
Biddle sees huge value in this for Rolls-Royce. “When the engine comes in for a service or overhaul, it could really shrink the turnaround times because I’d know everything there is to know about it from the capability in the field,” he says.
Data from sensors embedded in machines or products could be used to create a virtual environment depicting certain parts of a factory. By bringing additional data into the picture, an augmented reality could be created that would allow physical and virtual worlds to join. This augmented reality could be used for virtual prototyping.
GE’s Härtel says it may take a while before we notice we’ve been through a revolution: “But in 10 years’ time when we talk about it, there will be many things that have happened.”