A shift toward edge devices capable of conducting compute, storage and communication (as in fog computing scenarios) will possibly accelerate the discovery of early predictors and biomarkers in connected-health scenarios. “Manufacturers drowning in a torrent of data coming off of equipment like pumps or engines struggle to detect potential points of failure 24 hours before failures occur," explains Marcia Walker, manufacturing industry marketing consultant, SAS, which in April announced a Cisco-validated design for cloud analytics that is intended to move analysis closer to the edge, while preserving historical analysis, operational control and model development in the core data center.

"When you think of autonomous cars, airline engines, or pacemakers, it becomes obvious that the hundreds or thousands of variables that analytical models could communicate up to the cloud or over networks would cause load issues," says Walker.

It’s in these scenarios that analytical modeling may move to the equipment level, at the edge, so that only data related to alarms and actions traverse networks. "It is the need for speed that is driving many of the business leaders, who swore they would never allow data out of their facilities, to realize they will actually need to," notes Walker.

It's also the need to empower people, as it will be people, and not "things," that determine success. Whether data scientists or traditional business users, there is a growing need to enable fast, intelligent decisions, whether their tasks relate to fleet management or monetization of mobile usage, or other factors critical to desired business outcomes.

To empower business users of various expertise and knowledge, business leaders need architecture that not only captures enormous amounts of IoT data from sensors — regardless of the protocols — but that also is extremely user friendly so that people of varied expertise can readily assign different levels of importance to distinct types of data. This would simplify efforts to isolate the data most important to specific business outcomes. The complexity of the architecture to enable that is serious enough that CSPs can help enterprises at a high level, or on a case-by-case basis, determine which applications require instantaneous analysis and response, and where it makes sense to move analytics and/or data to the edge (as opposed to the cloud or central data warehouse), and then to convert data into insight most valuable to the business user or data scientist.

In other words, as the IoT magnifies the need for fast, intelligent decisions, “edge data” analytics platforms will be coveted by any company looking to build new revenue streams or business models through the IoT. Platforms will be sought, and either the NEPs will dominate, or CSPs will find ways to partner, leverage and monetize what they have and what enterprises cannot do alone, at IoT scale.

For example, Cisco has worked with IBM to integrate IBM’s Watson AI analytics into Cisco’s edge routers. And Israeli start-up Iguazio recently garnered attention in the press with its $33 million in funding this summer for its edge data analytics platform, as did startup MapR, which rolled out a new program for big data deployments.

Carpe Diem

While these platforms come into being, some CSPs are experimenting with how to partner in analytics, and in adtech and other areas of potential IoT growth and revenue. Verizon has been one of the more aggressive, evaluating how it can play an enablement role in the area of IoT analytics, becoming one of the investors in the aforementioned Iguazio real-time analytics platform through Verizon Ventures and launching at Mobile World Congress its Exponent portfolio of carrier-focused digital service platforms. It is meant to facilitate an “enabler” role for Verizon in growth areas like big data, AI, IoT, cloud, internet and media services by helping resource-constrained service providers and enterprises cut development cycles and capital expenditures with anend-to-end platform that delivers modular point solutions tailored to individual use cases.

“Verizon is building its intellectual property and working with Open Source to ostensibly white label solutions already running within Verizon so that carriers in emerging markets like Malaysia and Japan can build IPTV, media and other services and begin to capitalize on insights gathered from IoT for personalized value-add services,” explains Gartner’s Sicular. 

She notes that Verizon has recognized adtech as an area of possible growth in the IoT, building its Oath brand (the marriage of AOL and Yahoo) and partnering in new areas, as with its Uplynk Video Streaming service, which has been integrated into Stadium Digital’s All Access Platform, a digital loyalty and fan engagement platform.