Nmedia - Fotolia


Put microservices, cloud at heart of your IoT strategy

While many IT pros assume IoT is all about the network, other technologies, such as microservices and cloud, should play an equally important role.

The Internet of Things, which connects and enables communication between a vast network of sensors, devices and objects, is one of the most exciting IT concepts today. IoT proponents suggest it will revolutionize everything -- from self-driving cars to personal assistants. Vendors view IoT as an opportunity to add billions of new devices to the Internet and generate untold revenues.

Skeptics, however, see it differently. Some argue that IoT introduces billions of unknown and untrusted devices, while others expect sensor costs to explode and cripple deployment. Furthermore, some skeptics believe privacy and security regulations will nip the whole IoT concept in the bud.

Collectively, these arguments ask the fundamental question, what does IoT really mean? Most take the term at face value, believing it simply puts sensors and machines directly on the Internet. However, this model creates potential problems. If IoT sensors and control elements are on the Internet, they are subject to possible Internet attacks.

But it doesn't have to be that way. There are already billions of sensors and control elements deployed both in homes and businesses -- they're not just on the Internet.

Power-lines and direct wiring connect most of today's sensor and control elements to a single controller, which interprets and acts on the reported conditions. Recently, these controllers became Internet-accessible, and most smart-home installations use this model.

In a controller model, a homeowner or business installs sensors and control points linking to the controller. The control point then provides basic control functions and can also link to a mobile device for remote management. The controller must be online, but it includes security safeguards.

This controller model can address many IoT cost and security issues for users. However, other problems remain, such as data format issues, and locating and authenticating sensor information. Solving those issues could mean big changes for an organization's cloud environment.

Rethinking the IoT approach

Sensor and control devices accept a variety of data formats. Users connecting to an IoT sensor wouldn't know the sensor's data format unless all sensors had standard formats. And without knowing data formats, data interpretation and control are difficult.

Locating sensor information may not seem problematic, but anyone who drives through a busy city realizes there are many places to install sensors. All sensor data is situational, reflecting the conditions around it; so how do users find and interpret the necessary sensor data?

One solution is to convert IoT data into a database users can query through a microservice. If the authenticated data from IoT controllers is collected into a convenient big data format, then analytics can extract the data on-demand -- not only for the current period, but for past periods as well. The storage process can harmonize sensor formats into a common layout and index data by location or a variety of other factors that facilitate access and use.

This database-centric model turns IoT from a networking challenge to a cloud challenge. Users still have to collect IoT data, but also index and store it for easy access. Additionally, this model requires organizations to address IoT security at the cloud level, rather than the network level. Cloud assets growing underneath applications without direct application involvement -- as IoT assets do, since sensors are not part of user applications -- also requir e special planning to address data currency and to support synchronized analysis of multiple IoT sources. While current practices can likely address this, IoT application scale may prove challenging.

A database and microservice IoT approach also offers better support for privacy and public policy limits. Because query patterns are directly visible, IoT systems based on microservices and queries make it easier to detect attempts to track a person's location.

The mission of IoT is to connect assets providing valuable information and influencing important tasks. But an information flow from those assets is necessary. Adopting a database approach to IoT could provide all the information we need without the potential risks and costs.

About the author: Tom Nolle is president of CIMI Corp., a strategic consulting firm specializing in telecommunications and data communications since 1982.

Next Steps

Tackling IoT security challenges and risks

IoT in healthcare blanketed with uncertainty

PaaS providers create new offerings for IoT

IIC takes on Industrial IoT security

Dig Deeper on Cloud security tools