Yeah. This is my fourth decade in the information technology business. I was involved at the very early days in starting Oracle Corporation with Larry Ellison and Bob Miner, in building the relational database business.
My second company was a company that I founded in 1993, called Siebel Systems. Siebel was about the application of information technology and communication technology to the value chains associated with sales, marketing, and customer service. We largely invented the market that is known today as "CRM" or "customer engagement".
That company merged with Oracle in 2006, and in a very good outcome for Oracle, and for the company. Since 2006, we've been focused on making a contribution to the energy dialogue.
This infrastructure is undergoing an upgrade this decade, so that all of the devices in the infrastructure are being sensored in a manner that they're becoming remotely machine addressable. These include, one of the most common examples, which most people are familiar, is the smart meter. It also includes the thermostat, the variable speed fan in Walmart, the smart meter, the transformers, the substations, the vibration sensors on nuclear reactors. They're all being sensored to provide signals in real time.
Worldwide, the investment in this technology upgrade of this decade is two trillion dollars. This will be the largest, and most complex machine ever built. What we have done, is we've effectively been building the operating system for the smart grid.
Yeah. The datasets are staggering. If we give an example of Enel, which is our utility based in Rome, Enel operates 67 million meters in 40 countries. It's a utility roughly the size of the U.S. market. There, we're working on applications associated with revenue protection, which is identifying revenue theft, predictive maintenance, which is identifying devices that are going to fail before they fail. The economic and social benefits of that are obvious.
This application involves, for Italy alone, where they have 32 million smart meters, this involves the aggregation of seven trillion rows of data into a 700 terabyte cloud image, that grows at the rate of 300 gigabytes a day, and we're processing transactions currently at the rate of 1.5 million transactions per second, which is, by any standards, a record. We do this in large cloud-based, elastic cloud infrastructures.
The economic benefit of that, we're talking about increasing the safety, increasing the reliability, lowering the environmental impact of the processes in Italy and Spain, where the economic benefit, in Italy alone, is on the order of 350 million euros a year.
Predictive maintenance, revenue protection, the economic benefit of what we do is on the order of 300 dollars, per meter, per year to the utility operator. If you have 60 or 70 million meters, this is significantly non-zero. The economic benefit of what we afford Enel, is on the order of six billion euros a year in recurring economic benefit. That's not a very hard sell. The economic benefit that we afford Exelon is on the order of 2.7 billion dollars a year in recurring economic benefit.
I've been in the information technology business for almost as long as the information technology business has been around. I have never been involved with an application in information technology where the economic benefits, the social benefits, and the environmental benefits were so obvious.
Yeah. We have an analogous value chain in, and not a dissimilar value chain, in oil and gas. The difference is, that value chain is already almost fully sensored. They're much further advanced in the sensoring of the wells, the pumps, the lifting systems, the towed arrays, all of it. The data that are available are staggering, and the rates of data acquisition are breathtaking.
These data tend to be stored in discrete, siloed information systems for geophysical systems, or well production, or temperature, or rotational velocity, or whatever the physics are that are measured. We're able to, again, aggregate all of these data into a unified, federated image, and then analyze these data at machine speed to be able to ... let's take an example of predictive maintenance, at any point in time, eight percent of production facilities in the world are down, because of a device failure.
If we can tell an operator what devices are most likely to fail next, and why they're going to fail, they can either fix them, or replace them before they fail. The economic benefit of what we afford Royal Dutch Shell is on the order of five billion dollars a year recurring economic benefit.
When we get into oil and gas, we're dealing with very smart, very sophisticated people who are dealing with leading edge technology. These sensor networks are in place, and these information systems that they have are very, very rich. That's an enormous opportunity.
What we do is aggregate the data from these discrete systems, and be able to handle the influx of all these sensors to make sense out of it. The utility industry is not known to be leading in the information technology industry. The oil and gas industry is. They are absolutely ready for this and can perhaps realize the economic benefit.