“This is really important for us because the IoT is early stage,” said Steve Stover, VP of Marketing and Product Management at Predixion. “It is also complex, and there are a lot of moving parts. Because we aren’t trying to be all things to all people, we need to partner in this market.”
Predixion, which was founded in 2009 and first brought its cloud-based predictive analytics to market in 2010, got its initial traction in the health care space.
“That vertical provided us with a lot of opportunity to solve problems, then we expanded into more horizontal plays around sales, marketing and fraud,” Stover said.
Their move into the Internet of Things really dates from their Series C financial round in 2013, when two new strategic investors – GE Ventures and Accenture – invested specifically because of the perceived suitability of their technology to the IoT.
“We don’t try to sell head to head against IBM and other large vendors in the space and say our technology is just better,” Stover said. Their Predixion Insight cloud-based advanced analytics platform has patent-pending Machine Learning Semantic Model technology, which provides “predict anywhere” flexibility that Stover said was ideally suited to the IoT. It lets advanced analytics packages be embedded into a variety of production environments, including directly onto connected or disconnected devices, in addition to applications, databases, and complex event processing engines.
“This technology is well suited to push predictive analytics out to the edge, where there is limited connectivity mode,” Stover noted.
He cited as an example a customer in the smart mining space, which uses machines that have sensors which connect and gather data, but which have little connectivity.
“Because field operations were disconnected from the cloud, with their old technology, it took them two weeks to find out a machine had broken down, which impacted the work of other machines,” Stover said. “We can operate on a very small footprint, so action can be taken right in the field without data having to be moved into the cloud. It also allows for a much faster response than waiting for data to travel through the IoT architecture.”
Stover stated that throughout Predixion’s history, they had been very selective about entering into partnerships with other vendors.
“We had a handful of very strategic focused partners, who were not there just to put logos on the slide,” he said. “But because the IoT is so complex, we are now working with more partners for better access to the market.”
Wind River, a subsidiary of Intel, has a Helix Device Cloud solution designed specifically for the IoT.
“The Wind River Helix Device Cloud provides management infrastructure to manage the connected devices of the IoT,” Stover said. “We work with that to move our software out to the edge, and gain access to the data. We also provide the provide ability to build applications on top of the Helix Device Cloud, that can do things like understand the operational status of an oil well.” The solution, which can be deployed on-device, on-gateway, or in the cloud, will cut customer maintenance costs, better use their resources, and better protect against equipment failure.
“This goes beyond operational command and control,” he added. “Being able to build a predictive maintenance application to predict the likelihood of failure will allow a company to shut down an oil well on the brink of failure before things go red and it’s too late. They can also schedule how to best maintain assets, by effectively dispatching technicians in areas like oil and gas which have dispersed locations.”
Predixion goes to market through a multi-faceted channel strategy with includes direct sales, “sell with” partners like Wind River, partners like Accenture, in whose platform they are embedded, and reseller partners.
“The channel opportunity with the IoT is really about moving the needle for businesses who generate a lot of data, but who aren’t getting the value from it that they could,” Stover said. “Many partners, because of the dynamics of the early stage market, are looking to differentiate what they do, in terms of helping companies who are connected do more with their data.”
Opportunities vary significantly across verticals, because the level of IoT maturity varies significantly between them.
“Because the IoT came out of M2M, some industries like manufacturing where M2M is strong are more advanced than others,” Stover said. “They are moving from just collecting data to wanting to do things that are more meaningful with it. That’s a lot of the near term opportunity. In other markets, like health care, people still trying to figure out how to use the potential, and it is moving at a different and slower timeline. IoT might still be early stages in three years for those kinds of verticals.”