By Danny Pal in Computing
When your business supplies over 30 per cent of the planet’s power and produces engines for everything from cars to aircraft, it’s important to ensure that both the means of production and the products themselves operate as efficiently as possible.
That’s why General Electric (GE) – the fourth largest company in the world – has invested heavily in the Internet of Things and analytics technologies. As Bill Ruh, GE VP of global software services, told Computing: “It’s about getting better outcomes out of the machines we produce … the most profitable thing we can get to is zero unscheduled downtime.
“If you can make sure your power generation doesn’t go down, you don’t have brownouts, you can change the world and substantially improve industry.”
GE Aviation, the world’s second largest engine manufacturer, is using Internet of Things (IoT) technology to manage and repair jet engines pre-emptively, detecting tiny faults before they have a chance to grow into big ones.
“We’re finding that we can drive huge productivity increases in our teams because we can look at historical data across the fleet and begin to identify problems before they really happen,” Ruh explained
“We’ve had huge productivity increases because we’re automating how we think about maintaining an aircraft engine,” he continued. “We’re using big data analytics to help drive the maintenance schedule.”
By using sensors to collect engine data, GE is able to perform short bursts of maintenance much earlier on, meaning that over an entire lifecycle the engine spends less time in repairs.
Ruh likened it to getting a car repaired, but then discovering something hadn’t been spotted and having to return it to the garage shortly afterwards.
“The thing you hate when taking your car in is when in two months you have to take it back in. If they’d fixed it the first time, then they’d save you money and time,” he said.
“We can do that now. This is going to drive satisfaction rates and time on wing.”
But jet engine maintenance isn’t the only area where GE has harnessed the power of the IoT and big data analytics. Another is what the company calls its “Brilliant Factories Initiative”, which, according to Ruh, is “rethinking what a factory is”.
“Analytics in a factory has probably been underutilised, the data is mostly thrown away,” he explained.
“But we’re not doing that. We’re keeping every piece of data about how the machines operate and we’re using that to continue to identify types of analytics that could drive efficiency in our factory.”
Ruh added that GE is aiming for a “huge” 40 per cent improvement in factory efficiency.
‘If I could do analytics without IoT, I would’
But while connected devices and the Internet of Things are helping to drive improvements, for Ruh, they’re not the most interesting part of factory set-up; that would be the data and what can be mined from it.
“If I could do the analytics without IoT, I would, but I can’t because I need the machine data and machines are chatty. So for us, the IoT is necessary but it’s not the most interesting part. The analytics, the insight you gain: that’s where the value is. It’s just that you need the data in order to gain the insight,” he said.
For a company the size of GE, there’s a lot of data to gain insight from.
“Our current jet aircraft engines produce one terabyte of data per flight,” Ruh said, to illustrate the scale of GE’s data trove. “On average an airline is doing anywhere from five to 10 flights a day, so that’s five to 10 terabytes per plane, so when you’re talking about 20,000 planes in the air you’re talking about an enormous amount of data per day.”
Apart from the sheer volume of data, Ruh estimated that the firm analyses 50 million variables from 10 million sensors. These are the sorts of numbers that most industries could not conceive of managing. So when is IoT going to become useful to the mainstream?
Ruh believes that IoT going mainstream may be three to five years away. However, he warned that “if you’re not starting today you won’t be mainstream”. It is especially challenging, he argued, because operations teams rather than IT teams are the ones leading the change.
“The people who run power plants? They’re working with IT, don’t get me wrong, but they’re putting their own operations structure in place that is parallel to the IT structure,” he said.
“They’re already putting sensors in power plants, airlines and oil rigs. The number of sensors, the amount of data and the collection of that is already there,” Ruh continued. The trouble is, he added, the operations teams tend not to link this to analytics, which is where IT needs to get involved.
“The magic has to occur when the IT and the operations teams come together and that’s really where in five years the mainstream will be driven, when these guys are working in tandem,” he said. “GE does that today, but it took us four years to get there and it’s going to take other organisations that long.”
Despite being relatively advanced in its use of IoT and big data processing, GE isn’t resting on its laurels. Rather, it is exploring new areas to deploy the technology, including power generation.
“We’re changing the way we do remote monitoring and diagnostics. GE generates over 30 per cent of electricity in the world on our machines. So we monitor those machines remotely, then we try to proactively understand their behaviour and how to fix them,” Ruh said.
Ultimately, as with the jet engines, use of this technology will allow predictive maintenance, “telling our customers what they should be doing as opposed to what has happened”. This approach represents “the cornerstone of modernisation”, for GE.
‘An R2-D2 for every field worker’
But sensors are just the start. Ruh described how GE is looking to exploit the power of internet-connected robots to carry out “dirty, dull and dangerous tasks”. The firm has already experimented with this in the railway industry.
“We’re testing robotic rail inspectors that will go along and look for problems in rail yards, looking to see if there’s anything broken,” said Ruh. “The nice thing about that is they work 24 hours a day and they can work in the dark as well.”
GE called its first robotic inspector “The Guardian”, because as Ruh pointed out, it’s there to help people, not replace them.
“It’s something which actually works with a human,” he said, adding that GE got inspiration for this model from the film Star Wars.
“Most people think robotics is separate from humans but I look at something like Star Wars, where the robots weren’t there to replace humans, but to help them. So the question is: how do you create an R2-D2 for every field worker?”
Ruh went on to describe how this could raise some interesting questions in the future: as robots get more intelligent people could potentially start treating them on a more equal level.
“People are treating their robots like pets or members of the family. People are giving robot vacuum cleaners names, but not only that – and I think this is strange – but people take them on vacation,” he said.
“The fact is that we’re now seeing the Turing Test being passed every day, with people having no idea that they’re talking to a computer,” Ruh said. By extension, if the trend continues, robots will become widely accepted in homes and workplaces.
“Once robots take on human qualities – and I think that’s what’s going to happen -we’re going to find these things are playing a role in our lives as part of the family,” he concluded.