<< back Page 2 of 2

The Internet of Things - A Huge Opportunity, Wrapped in Risk

In-Depth Insights

Through advanced analytics, data streaming in from IoT can be mashed up with other data sources to provide in-depth insights. For example, Openshaw and his team illustrated that “a packaged foods distributor might mash point-of-sale data with weather forecasts, truck locations, and production data to predict stock-outs, reroute distribution, and reduce stale inventory write-offs.” Sensors within production equipment also can provide insights such as predictive maintenance.

Availability and Performance

IoT may offer the ultimate failover framework, as the steady stream of IoT data from multiple sources ensures that data is always available, even in the event one source fails. “Companies such as Royal Flying Doctor Service use IoT and big data to guarantee the availability of critical medflight services and medicines,” Splunk’s Gilmore illustrated. “Established companies like New York Air Brake use the IoT and big data to improve the performance of freight trains and the engineers that operate them—to the tune of a billion dollars in savings.”

Asset Tracking

“For enterprises that have many assets or high-value assets deployed across multiple locations, operational performance depends on balancing resource utilization, rapidly redeploying resources, and keeping assets in service,” Openshaw and his co-authors said. “Automatic tracking can eliminate the need for repetitive human labor, improving ease of use, removing human error, improving asset management, and making decision making simpler. For instance, shipping and logistics companies can precisely track the location of shipping pallets that carry expensive inventory—whether high margin or large volumes—through continuous monitoring of pallet condition, location, and movement via wireless relay.”

Real-Time Capabilities

Leveraging IoT provides data managers and consumers with real-time views on events or trends affecting businesses or individuals. “Healthcare and fitness devices like Fitbit and Jawbone are giving consumers real-time reports that can be used to improve their overall health,” Patel said. “Cities benefit by providing real-time information around electricity and water usage to help drive behavior change among residents and businesses and optimize consumption.” The potential of wearable devices goes beyond personal health monitoring, Openshaw and his co-authors pointed out. These devices enable companies “to digitize employees, which may include tracking and providing feedback on behaviors that affect wellness or job effectiveness, and could improve productivity across the workforce with the right supporting systems.”

Revenue Growth and Innovation

The Deloitte team sees the potential for IoT to facilitate “opening up new markets, providing better products and services through greater insight and management of product and customer lifecycles, and enabling relationships rather than transactions. This technology allows for greater visibility and control of the enterprise itself but also allows companies to understand and begin to shape their relationships with customers, business partners, and their own workforce.”

While the opportunities are clear, there are also many risks that need to be overcome to realize these advantages. First and foremost, there is the security exposure incurred as enterprises open up their systems to countless smaller devices across global networks. “Connectivity without consideration for security or privacy is such a huge risk that the companies which fail to take a proactive approach to these challenges may not survive,” said Gilmore. However, he added, many of the security platforms and techniques necessary are already in place or available. “IoT projects need to repurpose the security tools and principles already built for internet and mobile, and adopt new technologies to fill the gaps—this isn’t a net new challenge and may not need a net new solution.”

IoT may offer the ultimate failover framework, as the steady stream of IoT data from multiple sources ensures that data is always available, even in the event one source fails.

The ancillary concerns that come with data security—such as data ownership and transparency—are also amplified with IoT. “Important considerations include a clear definition on data ownership, sufficiently trained consumers on proper data use and sharing protocols, transparency that allows users to control who has access to their data, privacy and security guidelines and IoT governance bodies and consortiums,” said Singh. “A layered security approach is not sufficient for IoT; you need a holistic systems-based approach. Threat modeling and modeling of data, access, and control [have] to be looked at fresh. The conventional IT application and IT architecture is not relevant in IoT.”

Ultimately, IoT data governance requirements will move authority and stewardship out of the data management area and into the enterprise in general. “Traditionally, IT has been handling the data and systems, but with IoT impacting products, customers, and services, the data is more engineering- and product-related,” Singh explained. “Consequently, existing big data practices will not be applicable, resulting in a transformation in data management processes.”

Performance slowdowns pose another risk factor as enterprises move to IoT. In many cases, today’s enterprise systems are simply not ready to take on the surge of data and analytics that IoT makes possible. “Most of the big data technologies do a lot of batch processing,” said Patel. “This does not work in IoT implementation. Vast amounts of data—mostly as small snippets—are generated by IoT devices. This data needs to be available through real-time searches that provide insight into the stream of data coming in, which in turn helps make smarter decisions either by machines or humans.”

The ancillary concerns that come with data security—such as data owner­ship and transparency—are ampli­fied with IoT.

The tsunami of IoT data will bring “challenges in data transfer, handling, storing, processing, and visualization,” said Singh. “The wide variety of IoT solutions across the industry will also result in disparate data types such as time series, structured, unstructured, binary image files and text files. IoT implementations will also spur new data types, which will impose a challenge on the current big data technology in ingesting, storing, and processing.”

Ultimately, enterprises will need to step up their data management capabilities, and no organization will be able to go this route alone. IoT has interconnected layers that need to work together, including embedded systems, wireless sensor networks, control systems, home and building automation, and many others, said Singh. “No one single company or organization will be able to create a business model around IoT technologies on its own. It will depend on a dynamic ecosystem of partners to keep up with the continuous evolution in technologies and in the market.”

Ultimately, organizations may need to turn to wider arrays of solutions to be able to handle massive amounts of IoT data, Gilmore agreed. “IoT implementations fail when applications or technologies are stretched beyond their core competencies,” he said. “The most successful IoT implementations take advantage of ecosystems of vendors and applications that bring together fast time to value, ease of use, and a full stack of complementary features.” Many of the solutions are coming from the open source community, Singh also pointed out.

“Many of the big data best practices still apply, but IoT will challenge assumptions and demand new flexibility and capability,” said Gilmore. “There will certainly be a move to cloud or hybrid architectures as IoT data’s point of origin is likely to be outside the data center. There will be challenges due to diversity in both devices and stakeholders. There is a serious risk of creating big data silos by approaching IoT implementations with a project approach —where each project is built as a science project and survives only as a standalone environment.” IoT could and should provide the opportunity in which the data fabric really gets to shine, he noted.

<< back Page 2 of 2


Subscribe to Big Data Quarterly E-Edition