The Internet of Things (IoT) has exploded onto the scene, bringing us smart phones that track health data, drones that monitor wildfires and onesies that alert parents when their baby starts to move. A digital system where people, objects, and networks communicate and interact in entirely new ways, IoT is viewed as a major paradigm shift that will transform how we live and work. But is IoT really all it’s cracked up to be? And how will it affect the computing landscape that makes it possible?
Not Just for Consumers Anymore
According to a report by futurist and CTO Dave Evans, an average of 127 new things are connected to the Internet every second. While IoT has initially focused on consumer applications, industrial IoT is on the way. Clean tech businesses are pulling data from connected windmills, for example, with companies using genomic sensors or aggregating dispersed data for industrial purposes. Even government is getting into the act: the Homeland Security Department announced last year that it’s exploring wearable equipment for emergency first responders.
It’s a Brave New Computing World
Behind the emergence of IoT are major structural issues about information governance – especially as apps are moved to cloud. Connected devices consume and generate information that requires backup, recovery and management. To accomplish this, IoT requires a new, holistic approach to the data residing in endpoints, data centers and the cloud. Yet unresolved questions remain about data processing, backup and intelligence.
What Should I Do with All that Data?
An enormous challenge of IoT is that many organizations jumping on the bandwagon don’t yet realize they are now Big Data companies that have to process and manage massive data sets, including personal information and passwords. Currently, CTOs allocate engineering resources to this effort, but they’ll eventually want ready-made tools to manage workflows. Unless the people and companies driving IoT can get their arms around Big Data, privacy and security concerns will thwart mass adoption.
IoT – which often collects information unbeknownst to users – raises a host of privacy, security and liability issues. Yet authors of a report issued by the Institute for Critical Infrastructure Technology (CIT) note that IoT often lacks any form of security, representing “practically an infinite attack surface” for cybercriminals. The risks are real: according to Marc Rotenberg of the Electronic Privacy Information Center, “If you think you’ve got a cybersecurity problem now, wait for the cold winter day when a hacker halfway around the world turns down the thermostat on 100,000 homes in Washington D.C.” As a result, IoT solutions will require bullet-proof functionality to protect sensitive commercial information and safeguard users’ personal data.
The Edge and the Cloud Duke it Out
Traditionally, data has been collected from endpoints and sent to the home server or data center for processing. Because this approach is ineffective for handling massive data sets, organizations are increasingly using edge computing to maintain and process IoT data locally. At the same time, it’s routine for sensor and other IoT-generated data to be moved to the cloud. Since there’s little agreement on which approach is best for processing, IoT employs a mix of edge and cloud computing, with data management and protection strategies required for both.
Global Dedupe Lends a Hand
Global dedupe has proven key for transmitting and managing IoT data. By understanding data patterns that are common to devices, global dedupe transmits only what’s uncommon, thereby reducing enterprise bandwidth transmission by about 90 percent. While global dedupe uses the GPU to achieve exceedingly fast fingerprint processing, its intelligent use of CPU cycles can also help minimize user disruption.
Data Classification Moves Front and Center
IoT data classification has become a hot topic given its ability to optimize information backups, governance and workflows. By classifying data based on variables of its choosing, organizations can optimize data management (e.g. here’s information to be disposed of versus stored and retrieved later on). Data identification is currently a manual process but this will change, as auto-classification evolves over the next few years.
IoT is here and the sooner organizations adapt to the new normal, the more likely they are to benefit from the enormous opportunities that have opened up. Fortunately, solution providers are busy developing customized tools and systems for IoT. And better yet, organizations that understand the dynamics of IoT can keep their ear to the ground and quickly embrace new approaches so they can thrive.
View the eWEEK slideshow looking at seven ways IoT will impact the computing landscape.