Kevin Ashton coined the term “Internet of Things” [IoT] in 1999. He built the concept primarily on his experience and expectations for Radio Frequency Identification [RFID] adoption. Since then, the IoT concept has grown to encompass any device that uses any type of Internet connectivity. Let's take a look at some of the myths surrounding the IoT and take a look at why it may never reach its potential.
Not all Connectivity is the Same
Many predictions have come forward on the number of connected devices and their economic impact.
- GSMA announces the Business Impact of Connected Devices could be worth $4.5 trillion in 2020
- Cisco: The Internet of Everything Value Index
- Ericsson: The Internet of Things Comic Book
- Microsoft StreamInsight - Building the Internet of Things
Some of these consider only smartphones, tablets and similar devices using wireless networks, and others treat a hand tool with one sensor and a vehicle with hundreds of sensors equally as a device; some only consider wireless connectivity while others consider both wired and wireless connections -- these are difficult to compare.
They are also all somewhat misleading when trying to consider the impact of IoT on data management and analytics [DMA], both because of the disparity in number of sensors per device, and in ignoring the different impact of casually connected devices versus continually connected devices; for example, a fitness tracker that uploads data via Bluetooth Smart to a smartphone to the cloud only when the user touches the “upload” icon versus an image sensor that is constantly streaming live video of a city street.
The Role of Data Management and Analytics
At this point, we need to ask “What is the IoT?” The Internet grew from ARPAnet, expanding the ability of humans to interact with other humans and with machines [H2H and H2M], primarily for communication and transaction processing.
The IoT is more about Machine-to-Machine and, ultimately, Machine-to-Human [M2M and M2H] interactions. The true impact of the IoT will be from Internet of Things Analytics [IoTA], also known as Connected Analytics. But not just from analytics alone. There are social, economic and political imperatives causing solution spaces to form around IoT, M2M and M2H data. These solution spaces are intersecting to form Sensor Analytics Ecosystems [SAE].
There are many initiatives driving the IoT, and all center around DMA:
- Cisco: Internet of Everything
- Informatica: Vibe Data Streaming and Virtual Data Machines [VDMs]
- Teradata: Analytics as the Hub for Monetizing the Internet of Things
- Wipro: Connected Analytics
- IBM: SmarterPlanet and other “Smarter” initiatives
- Salesforce.com: Internet of Customers
- Oracle: Device-to-DataCenter, and J2ME for IoT intiatives
- GE: Industrial Internet
- Metaio: Augmented Cities
- Government: Green IT, Smart Cities, Smart Government, Traffic Management, Parking and Infrastructure, and many, many more
- Individuals: Quantified Self, Personalized Medicine, Home Automation, added convenience and value in shopping, service and support
Will the Internet of Things Live Up to Its Potential?
The IoT has strong potential to change how we make decisions, but it may never reach that potential. Currently, most connected devices feed silos of data and are single purpose; it is difficult to get data out of one silo, or use it in another. There is no commonly accepted standard for data exchange among these silos. However, there are evolving models, such as standard TCP/IP, SoA and ReST, Messaging Queue for Telemetry Transport [MQTT] and the AllSeen Alliance.
In addition, the major corporate and government initiatives each view the IoT in different ways, and these initiatives are often confused and seen as identical, when they’re not. I’ve written before of these problems, and of the solution that I believe will allow the IoT to reach its potential: Sensor Analytics Ecosystems [SAE]. These SAE will evolve around specific solution spaces. How these SAE evolve, and their impact on DMA, depend upon
- Intersections among the solution spaces
- Ownership of data
- Pivoting privacy, transparency, security and convenience on the fulcrum of trust
- Marketplaces for raw data, cleansed and analyzed data, and insight from inferences and prediction built upon data
- Services within SAE to help individuals, businesses and governments adopt and successfully use connected DMA
Figure: Solution Spaces Intersect to Create SAE and Meet IoT Potential
How can we make an attempt to determine the impact upon DMA for individual practitioners, data centers, organizations small and large, private and public, and the DMA market?
- Count sensors and actuators, not devices, looking at data types, feedback loops and interfaces, starting with current connected and non-connected components to understand the impact within your organization.
- Consider the impact of casual connectivity needs versus continuous connectivity.
- Look at architectures: Elastic Cloud Provisioning, Cisco’s Fog Computing and similar edge computing/cloud computing architectures, and IoT specific architectures and platforms.
- Remember that while big data is a joke, the resulting data science programs are changing the face of DMA, Statistics and Computational Statistics -- but keep an eye towards bringing data science results into production for use by all types of decision makers.
- Don’t worry about defining IoT, focus on the solution spaces of importance to you, and how your approach to IoT DMA will make you a successful partner in SAE and data marketplaces.
- There are a multitude of data types and current analytic tools that best handle a subset of sensor data types.
- In addition to sensor data types, IoT data will require third-party data to fully, deeply understand this data and bring context for decision support.
- Many industries will go the way of the photocopier market: the equipment is marginalized, data from and services for that equipment are profitable.
- Take a System approach, not systems, and certainly not devices; treat interfaces as multipliers, be aware of events and faults for cause-consequence analysis.
- Understand the interfaces, interactions and interchanges of the SAE in which you will participate.
Title image by Albert Lozano (Shutterstock)
Editor's Note: Read more from this month's focus on the Internet of Things here.