Analytics/Big Data

Cisco Paints IoT Into the Big Data Picture

Network operators' interest in the Internet of Things (IoT) is less about the number of connections they can add to their network -- although those numbers are staggering -- and more about the data those connections bring with them.

The amount of data that machine-to-machine (M2M) devices generate is massively significant. It's data that consumers can use to change their behaviors, like how a FitBit may encourage you to walk more. And it's data that operators want to use to tailor their services and how they market to consumers.

Recognizing that the size and nature of the data is of a different scale and scope to that which operators have dealt with before, Cisco Systems Inc. (Nasdaq: CSCO) has built -- through its acquisition of Composite Software last year -- a way to ingest IoT (or Internet of Everything to Cisco) data, along with all the other disparate cloud and network data operators have into one consolidated view. (See Cisco Beefs Up Its Intercloud, Adds Telco Partners .)

The software has been available to network operators and enterprises for a while now, but with an update Cisco announced on Wednesday, users can now search and query data on their own without the need for intervention from the IT department, allowing them to make decisions faster.

Mike Flannagan, GM of Cisco's Data & Analytics Business Group, says that the update also includes a software development kit for startups to connect to its data virtualization product, which supports more than 150 data sources today.

Want to know more about telco analytics? Check out our analytics content channel here on Light Reading.

"The primary driver for this is to remove data silos and give enterprise applications and intelligence visibility to all the data across an enterprise," Flannagan says. "It's a problem in the data center. What we're seeing with IoE use cases and connecting so many sensors, is that so much of the sensor data doesn’t need to be sent back to data centers."

As an example, Cisco's oil and gas customers generate massive amounts of data on offshore platforms that would take weeks to transmit back to the network core. They need to know what information needs to be sent and what can be stored locally in real time.

Cisco provides its Data Virtualization software to enterprises and telecom service providers and, often, enterprises via telecom service providers. Flannagan says its biggest customer is a North American telecom provider, which it is helping break down silos across its business.

Data virtualization is complementary to what Cisco is calling Fog computing -- the idea of embedding intelligence from analytics in edge routing devices to make decisions locally. Both involve local processing, but data virtualization deals with querying the data that fog computing might produce. (See Cisco Puts a Fog Over IoT.)

For network operators to ever tap into the vast world of IoT data, breaking down organizational silos will be one of the first and hardest steps. That's why it's an area attracting a lot of interest from big data startups, established equipment and networking vendors and essentially any company with a view of connected data. Flannagan says Cisco can work with any of these companies to ingest data and display it in an aggregated way, whether the source is IoT, cloud, data center, big data or otherwise. (See Big Data Attracts Big Dollars, New Faces and That Big Data Sinking Feeling.)

"We hear from our customers and analysts that most data analysts spend 60% of their time just trying to get access to data they want to analyze," Flannagan says, adding that silos, complexity and IT projects that often take up to nine months are too blame. "The interesting thing for customers is to do analytics that really leverage all your data across the company and get new business insights and use that for better sales."

— Sarah Reedy, Senior Editor, Light Reading

danielcawrey 10/1/2014 | 9:23:08 PM
Re: Removal of Friction Essential I think what is key here is Cisco's ability to categorize the date and the ability to recognize when real-time information needs to be disseminated immediately. I see that as very useful. 

These IoT devices are going to generate so much data, some of it is bound to have very little value. The filtering is key – it seems that Cisco is ahead of the game here, which it usually is via its M&A activities. 
bosco_pcs 10/1/2014 | 2:17:40 PM
Re: Removal of Friction Essential @Sarah, agree with you totally about silo busting. And that hinges on how the organization is set up. Companies have spent a lot of time and resources to play around with the centralization/decentralization theme to connect the dots. While KISS still rules, sometimes you do need to be pragmatic knowing when to have an exception. At my former shop, the general contractor should reach out to additional stakeholders but the latter might have to pony up additional fundings. 
sarahthomas1011 10/1/2014 | 1:59:26 PM
Re: Removal of Friction Essential You both make good points. There's a reason operators employee BI specialists. It's not easy to understand and take action on all this data. I'm sure the answer has to be a mix of both central and edge data storage, IT and self-service. The important thing, to me, is being able to see across silos. You might not want all the data, but I do think you need to have a complete customer picture to make any decisions, especially if you're a triple or quad-play provider.
bosco_pcs 10/1/2014 | 1:13:41 PM
Re: Removal of Friction Essential Frictionless is a great buzzword these days but there is a case to be made against neutron bombing human intelligence. I used to work in BI & Analytics - don't mind if I do something like that again - and the struggle was to find an optimal point of collection of data.

Obviously, it is great to collect everything and have every bit of data at one's disposal but there is a finite budget on storage. On the other hand, it is a pain to miss a column here and there, irrespective of what model/methodology your shop uses. That is why you need seasoned analysts and IT pros, as all the stakeholders' proxy, to scope it out for a relative durable system.

Ultimately, the good old cost benefit analysis is as relevant as ever before

Phil_Britt 10/1/2014 | 11:00:28 AM
Removal of Friction Essential The removal of IT intervention is essential because of the inevitable bottleneck. Of course, if people try to do this on their own, do something incorrectly and then have to involve IT, it will take even longer (same reasoning as why I call a plumber rather than trying to do it myself). 
Sign In