olly - Fotolia
IT teams must confront IoT integration challenges to achieve not only IoT integration, but true IoT interoperability.
The IoT device plurality gives organizations the opportunity to apply IoT in a variety of ways and industries, but it also creates the IoT integration problems. IoT's use of proprietary OSes and device settings make IoT products difficult to use with other technology and systems.
True IoT interoperability means all IoT devices and services use the same standardized data elements. If every device and system use the same data elements, these elements can be stored in a single data repository that provides one uniform version of data in all departments. One data standard can ensure technological elements can work together and deliver more accurate results.
However, the real IoT world is far removed from the ideal of interoperability. The IoT industry has no widely adopted data standards that IoT vendors conform to, which makes integration -- not interoperability -- a more realistic goal for IT teams.
IoT devices might use different names and formats for the same data, but the expectation is that IT admins can integrate IoT data through APIs, which translate data names and formats into ones that a receiving application can use and understand. But IT admins can take additional steps to overcome IoT integration challenges.
Create an overall strategy for IoT integration
Early in the emerging IoT market, organizations began piecemeal implementations of IoT, such as installing IoT sensor systems that monitored temperatures and humidity in buildings. With initial technology limitations, organizations didn't necessarily put much thought into how to integrate IoT with other enterprise systems, so the systems quickly became data silos.
As organizations developed more IoT deployments, development teams began to understand the need to integrate IoT and its data with other aspects of their business strategy. The demand for an IoT integration strategy grew.
Today, any IoT integration strategy must at least address:
- whether IoT OSes and data are standard or nonstandard;
- what effect IoT has on corporate network traffic;
- who within the organization can use IoT data;
- how IoT conforms with enterprise security and governance policies; and
- how to easily integrate IoT with other enterprise technologies.
Organizations must ensure any IoT vendor in consideration checks all these boxes on any request for proposal (RFP) issued. As part of the RFP process, organizations should look for true interoperability on an IoT vendor's roadmap and ask IoT vendors if they have plans to work with NIST, the Open Connectivity Foundation or other organizations that advance IoT interoperability standards.
IT admins should flag and consider upgrades or replacement of any IoT products that can't easily integrate with other systems.
Don't settle for non-integration and demand vendor support
Unless an IoT deployment is so unique and invaluable to an organization that they simply can't do without it, never settle for an IoT product so proprietary or underdeveloped that it can't interface with the APIs of other IoT devices or technology.
If the selected IoT vendor flees after the sale, look for another vendor that will give the support the organization requires to integrate the previous vendor's product.
The best way to assure this support is to discuss present and future integration needs for the vendor's product upfront. Vendor integration support should also be defined and written into the contract before signed.
Ensure on-staff IoT expertise or get external support
IoT must be properly configured for performance and security, and it must be integrated with other systems.
Most organizations address technical IoT needs by training internal staff and augmenting staff with outside expertise, as needed. Often, organizations turn to IoT vendors for expertise, but they should also look to their legacy system vendors that have APIs and integration expertise for IoT.
Regardless of the vendor, an organization's end goal should be to transfer expertise to internal staff so they can become self-sufficient.
Another approach to IoT and system integration is to use the cloud, which includes PaaS providers that specialize in IoT integration. PaaS APIs and experts can assist with integrating IoT with other cloud-based systems and can ease the IoT integration workload. By hosting IoT and system apps in the cloud, organizations also ease the amount of bandwidth needed on in-house networks to support IoT traffic.
Check networks and bandwidth
Organizations should check the capacities and throughputs of the networks they plan to use to transport IoT data before they implement IoT. As part of this exercise, IT admins must evaluate what IoT data they plan to use and integrate and what data they can discard.
IoT indiscriminately generates data of all kinds, and not all of it is relevant. When organizations define in advance what data to use or discard, IT admins prevent networks from being overloaded with excess data traffic.