How businesses are benefiting from the technology behind CERN's Large Hadron Collider

The internet isn't the only mainstream technology to come out of CERN

Deep below the Franco-Swiss border, the 27 kilometre long Large Hadron Collider (LHC) operated by CERN, the European Organization for Nuclear Research, is one of mankind's most ambitious scientific projects.

The LHC generates one petabyte of data every single second as it smashes particles together at almost the speed of light in CERN's quest to find answers to some of the deepest questions in theoretical physics.

How was the universe created? What are the smallest particles in existence? These are just some of the fundamental questions about the universe physicists are looking to answer, with a budget of some £750m a year.

That might sound like a lot, but it works out at about the equivalent of two cups of coffee per year for every European taxpayer.

Those wondering whether even this modest outlay represents good value for money don't have to look very far, as physics professor, scientist and former CERN researcher Dr Michael Feindt explains during a tour of the CERN facilities near Geneva.

"The technology needed to do this is so complicated, it's a technology driver. One thing that you all know which came out of CERN was the World Wide Web, the HTTP protocol was invented here," he says, describing the system invented by Sir Tim Berners-Lee that has "completely changed the way of life for billions".

The internet was created because CERN scientists wanted an easier way to share and access data.

"The internet was invented because physicists wanted to collaborate and they wanted to have access to the data of colleagues without having to know which computer they needed to know a password for," Feindt explains.

But the internet isn't the only thing to come out of CERN that has provided lasting benefits to the wider world away from the arena of theoretical physics. Other technologies have emerged too, particularly in the area of big data and analytics.

The reason these technologies emerged is because even the best physicists in the world can't analyse every single thing that comes out of the Large Hadron Collider, because the sheer amount of data produced is too much to comb through.

"We have to reduce the data because otherwise, if we wrote everything to CD, it would mean a stack of CDs the size of the Eiffel Tower every second," says Feindt.

CERN's computer farms therefore use high-powered algorithms to eliminate the "uninteresting" data, a difficult task given events such as the Higgs Boson collision are so rare.

"Because there are 10 million collisions which are completely uninteresting to us, then there's one which is perhaps interesting," he says, adding how 30,000 computers work in parallel to filter out the "uninteresting" data.

Feindt founded Blue Yonder in 2008 so that organisations outside the research institute could use the same predictive analytics as CERN's scientists.

"We do at the company something very similar to what we do at CERN," Feindt tells Computing, adding that the Blue Yonder algorithm is now available to retailers, insurers and other businesses.

"It's the same algorithm; it runs on different data, the input data is different but in principle, it's the same [as CERN's]," he says.

For the algorithm, Feindt adds, it doesn't matter if it is being used to make predictions about energy particles in the Large Hadron Collider or predictions about stock for a retailer, they're both a selection of mathematical inputs that can be crunched to find an answer by extracting and analysing the relevant data.

According to Feindt, the Blue Yonder algorithm does this by using the same method CERN uses to identify the most relevant data for analysis and discarding that which is deemed uninteresting.

"First we know when we have many, many measurements for a single prediction. Our algorithm finds out itself whether each of the variables is a statistically relevant connection to the prediction target or not," he says.

The big data crunching techniques of CERN have already trickled down to the insurance industry.

"In insurance you can make a better prediction for each individual, what his risk profile is. You can better design their tariff," Feindt says, describing how the algorithm can allow an insurance firm to both take more calculated risks and provide cheaper services for their customers.

"If you want to earn more money then a classical business decision is to make the overall tariff more expensive. But then of course then your tariffs are worse for your customers and your competitors may have a cheaper option than you and attract them," he explains.

"With our algorithms, we can individualise it to each individual customer and make individual decisions on tariffs," Feindt continues, before describing how this benefits both the business and consumers.

"You can gain more profit - you have to pay less percentage of your tariff back to insure people," he says. "But at the same time, you get more new customers because for many the tariff will be better and cheaper."

The analytical techniques used to crunch data generated by the Large Hadron Collider can also provide benefits to retailers by ensuring that they can make accurate predictions about what levels of stock they need.

"You want to always have products and have your shelves full, but you don't want to throw away items, meat for example, because the shelf lifetime has gone," Feindt says. "But we can work out exactly the right amount at the right time in the right shop, depending on all sorts of influences like weather forecasts or whether it's the weekend or not and so on."