GE has announced that its Predix software platform will launch in a cloud implementation next year, transitioning the Predix data analytics and machine learning capabilities from private clouds into a GE-controlled cloud service. Beginning in the fourth quarter of this year, current GE users of Predix will begin migrating their deployments to the Predix Cloud, with the full platform available to external customers in 2016.
Predix will be using Pivotal’s Cloud Foundry to enable application development, deployment and operations. Pivotal is a joint venture between EMC, VMware and GE, and Cloud Foundry itself is an open source computing PaaS that was initially developed by VMware but is now solely owned by Pivotal. The PaaS software will be able to run on Amazon’s AWS servers, but GE says the actual deployments will be run on its own hardware – as a security measure that large industrial businesses will surely appreciate.
Predix was unveiled in 2013, and was built to store and analyze incoming data in real time, on third-party hardware – mostly Amazon. Now running on GE’s own hardware, focused on industrial deployments, such as power generation and jet engines, the idea behind the Predix system is to use analytics to spot potential efficiency savings, as well as carry out predictive maintenance – to spot a problem and solve it before it becomes a much bigger problem later on, such as a faulty fan turbine developing into a total jet engine blowout.
Sensors in the industrial equipment collect and relay data to the analytics platform, which essentially crunches numbers at massive scale to spot trends in the data that might indicate a problem – or simply chart performance by visualizing its findings on a dashboard or graph.
“A cloud built exclusively to capture and analyze machine data will make unforeseen problems and missed opportunities increasingly a complication of the past,” said Harel Kodesh, VP and GM of Predix at GE. “The Predix Cloud will unlock an industrial app economy that delivers more value to machines, fleets and factories – and enable a thriving developer community to collaborate and rapidly deploy industrial applications in a highly protected environment.”
With security as a key selling point, GE is hoping that its industrial-only cloud will begin drawing in customers who might already be using the general purpose cloud services from the likes of Amazon, Microsoft, IBM or Oracle. While its competitors will, of course, claim their systems are secure, GE seems to be banking on convincing interested parties that deploying their applications on the dedicated GE cloud is the safer option – although that does sound a little like an attacker’s paradise, should the locks fail and a way in be found.
In addition, GE seems to be charging a premium over those alternative cloud providers, on the basis that its security systems require more computational resources. Having a stricter set of entry requirements to its servers (called the “Gated Community approach”) and resulting greater insight into what is currently being executed on them should provide it with a more secure operating environment, but only time will tell to see if the approach stands up to scrutiny.
“Cloud computing has enabled incredible innovation across the consumer world. With Predix Cloud, GE is providing a new level of service and results across the industrial world,” said Jeffrey Immelt, CEO of GE. “A more digital hospital means better, faster healthcare. A more digital manufacturing plant means more products are made faster. A more digital oil company means better asset management and more productivity at every well. We look forward to partnering with our customers to develop customized solutions that will help transform their business.”
GE notes that it had $4bn in software revenues in 2014, and expects to achieve $6bn in 2015 as it continues to grow its software investments. That’s still a fraction of GE’s total 2014 revenue – which stood at $148bn, making it the fourth-largest company in the world.
But the shift to the XaaS business model has a number of benefits. For the customer, effectively subscribing to a service allows them to spool up resources to match demand – meaning that if demand spikes, it’s simply a matter of renting more servers. Similarly, it demand falls, the customer can simply turn down the number it rents, without being saddled with the capex bill of having idle privately-owned servers left unused. For companies like GE, selling to these customers appears to be increasingly lucrative, as more and more users shift their operations to a cloud-supported architecture.
TempoIQ and the space for the smaller players:
And on the topic of those customers, this week we spoke to Steve Subar, the CEO of TempoIQ, a PaaS vendor that he says competes at a different tier than the likes of GE. Subar said that there is enormous space for competition for companies like TempoIQ – which sells it services to businesses looking to add some element of cloud functionality to their operations.
Because of this space, Subar says that one of the hardest aspects of business is keeping up with demand, and that because of this, he is confident that another SAP will emerge from this space – to challenge those aforementioned incumbents.
TempoIQ’s portfolio consists of 4 tiers, that can be combined for the full effect or used in combination with existing assets: moving from the database (ConnectIQ), to the storage and reporting (DataIQ), to the streaming analytics (AnalyzeIQ) and finally the dashboards and visualization (ViewIQ).
The modular platform is hosted on Tempo’s own cloud servers, with APIs to bridge data to other applications if needed. Choosing to monetize the platform by way of data points processed per month, rather than total number of connected devices, Subar added that the company doesn’t currently have any patents – as the political climate in the US has cast some considerable uncertainty over the viability of software patents.
In terms of strategy, Subar sees TempoIQ’s offering as an inversion of the SaaS acronym – that the company provides professional services in software, rather than software as a professional service. To this end, TempoIQ offers the Agile Accelerator service (formerly the Test Drive), which is designed to quickly build and deploy an application in 4 weeks.
That’s not necessarily a 4-week period, and is often four one-week chunks, at the end of which the customer will have a Minimum Viable Product (MVP) – which can then be fleshed out to meet arising requirements. Subar says the company has very little churn.
Notable customers currently include Wilson, the maker of sports equipment, which came to TempoIQ needing an additional application to port data collected by its soon-to-be-launched smart basketball and consumer app. Wilson used Tempo to quickly analyze the data pulled from that app to gather insight into how to sell more basketballs – its core business. The sensors and consumer app had already been completed, but Subar said Wilson chose TempoIQ for its integration speed –a central attribute, and one evoked by TempoIQ’s name.
Wilson is an example of what Subar refers to as a Connected Product Maker, one of two types of customer in this space, with the other type a Connected Asset Owner or Operator. For companies like Wilson, TempoIQ’s value comes in deriving data from a process or product; essentially connecting something to a cloud application that wasn’t connected before.
Companies like Danfoss fall into the second camp that Subar outlined, as it already has a connected product but required an additional service. Danfoss builds and sells electric motors, the likes of which can be found in the cranes that power docks.
These motors already had around a hundred sensors, which managed primary functions such as RPM and position, but the system was essentially throwing this data away. After turning to TempoIQ, Danfoss now collects and uses this data to provide remote access control and predictive maintenance – all without changing the physical product.