Data Analytics

Expanding IoT Necessitates Better Data Storage and Management Solutions

As internet-connected devices create an ever-growing mountain of information, organizations will need to explore new ways to store and make sense of the data.

BizTech Staff

Big Data is getting even bigger.

Experts predict that there may be as many as 200 billion devices in the Internet of Things (IoT) by 2020. All of these devices will be collecting, aggregating and sharing a massive volume of data.

As organizations plot the infrastructures they’ll need to support their IoT efforts, they can’t look only at requirements for today. They must plan several years ahead so their investments don’t go to waste when their data exceeds their storage solutions.

Storing IoT Data in the Cloud

It’s nearly impossible to gauge an organization’s future storage needs considering the limitless possibilities of IoT. So it’s important to find scalable, flexible storage solutions that can grow along with an IoT deployment.

Since various organizations have different resources and needs, IT leaders should consider a variety of storage options, including disk, cloud, flash and hybrid.

Cloud-based storage, which organizations can easily scale up or down, makes sense. Instead of housing the infrastructure in an on-premises data center, organizations can purchase storage from cloud-based providers based on their current needs. If those needs change, they can purchase more or less storage.

This Infrastructure-as-a-Service model cuts back on IT costs. Instead of buying more than they need to cover themselves during peak periods, organizations can scale up and down to meet changing demands.

As data flows increase, not only must organizations have enough space to house information, but they must also make sure data is easily accessible. IoT data comes from vastly different environments in different formats using different languages. Users, in turn, face a serious challenge in compiling these disparities into a common language and storage solution so it can be analyzed as a whole.

Software from vendors such as Splunk can help organizations make sense of the data they collect. Splunk and the developer community that has emerged around the company have created numerous applications and add-ons that simplify the collection of IoT data from disparate systems.

Data Analysis and Business Intelligence

Gaining value from IoT isn’t only about putting sensors on objects and connecting them to the internet. That strategy can backfire. If organizations invest too heavily in connecting objects and gathering data without making sense of the data, they can wind up wasting time and money.

While many organizations recognize the potential of IoT, they struggle with how to pull value from it. Even organizations that have a clear vision of how IoT can help them must overcome hurdles to get there. They need to determine how to automate the collection of data, combine different types of data — structured and unstructured — from various sources with different protocols, analyze that data, and implement it into their processes.

Data produced, collected and analyzed by IoT can transform organizations and help them make faster, better decisions with significant returns. In fact, an economic analysis by Cisco Consulting Services projects that IoT will generate $8 trillion in potential bottom-line value within the next decade through innovation and revenue, asset utilization, supply chain and logistics, employee productivity improvements and enhanced customer experience.

One of the daunting aspects of IoT for many IT leaders is starting over with new technology investments, but in many cases, that’s not necessary. In fact, integrating IoT initiatives with existing data analytics tools, such as enterprise resource planning (ERP) and customer relationship management (CRM) systems, can improve analysis.

Bringing all the data from ERP, CRM and other IT systems together with thousands of sensors on operational equipment can overload a network. But IoT sensors can be configured to communicate to the network only when there’s a notable change. For instance, a tire sensor wouldn’t clog a network with continuous data. Rather, it would process that data on the edge of the network (where it’s generated) and send updates to the network center (the data center or cloud) only when a change in pressure is detected. This concept of processing data at the network edge is, in some cases, referred to as fog computing.

Another critical aspect to IoT analysis is database management. Normalization is the process of bringing the varied sources and types of data together in a central location with a common format so it can be searched and analyzed. A strong database management system can help with this process.

Plenty of organizations are already reaping the rewards of IoT. For instance, the mining firm Dundee Precious Metals installed a wireless network covering 50 kilometers of tunnels, put IP-enabled sensors on everything from lights and fans to vehicles and video surveillance cameras, and equipped workers with IP-enabled phones, tablets and hats with radio frequency identification (RFID) tags.

The IoT initiative gave managers better visibility into what was happening underground, and it improved communication with miners, resulting in smoother operations. That translated into safer work conditions, a 400 percent increase in production and reduced maintenance and energy costs.

Future IoT applications are out there today, just waiting to be discovered from within the data.