In my last article, I indicated that an iPaaS (integration platform as a service) can be a good fit for IoT-related integration projects based on the fact that most IoT services have an API.
While I continue holding to this position, there is another perspective on the connection between IoT and iPaaS uttered by Gartner in their research note “IoT Drives New Integration Challenges and Best Practices” published earlier this year.
In fact, this perspective made me re-think the particular use cases when an iPaaS alone can be enough and when it should become a part of a complex mechanism. So, let’s review this perspective together.
IoT platforms alone are only “good enough” for the initial stages of IoT projects
I want to make it clear straightaway — I won’t give away too much of the research note, because you definitely need to read it completely if IoT integration projects are your topic. So, I’m not going to get into detail about where is the place of an iPaaS in IoT integration projects and why project implementers should turn to an iPaaS in the first place. But what I will reveal is that an iPaaS is believed to be the right solution to enter the scene when it comes to scaling up IoT projects and delivering more complex integrations.
This actually makes sense big time: IoT devices (Gartner uses even a broader term “IoT endpoints”) are not independent, stand-alone solutions. They need to be managed, data coming from them needs to be accumulated, analyzed, distributed and consumed in a proper way, IoT streams need to be monitored. All this can be achieved by implementing an IoT platform — “a software suite or cloud service that facilitates operations involving IoT endpoints as well as cloud and enterprise resources”.
Moving further, most IoT platforms already have specific, albeit limited embedded integration capabilities. While these capabilities are most certainly insufficient for complex IoT integration projects, they are good enough to start such projects in the first place.
When initiating an IoT project the main focus lies on how to bring it to life at all
Because let’s face it — we hear and read the buzzwords “IoT” and “Internet of Things” everywhere, but very few companies really know at this point how to tackle this topic to begin with, let alone carry out full-fledged IoT implementation projects. According to one of the latest Forrester’s global researches [PDF], only 21% of enterprises have implemented IoT, and I bet that the majority of them are tech companies, who know their way around innovation and technology anyway. Our own humble experience shows us that at this point, enterprises are eager to embrace the Internet of Things, but do not have a clear IoT strategy yet in place.
So, IoT projects are a very new phenomenon, and there is not much accumulated experience in this respect. It is only natural that the first tentative attempts at implementing IoT projects focus on grasping and embracing the innovative technology itself. It is only later that project implementers look for the way to scale projects in complexity, usually when they have already made themselves more familiar with IoT and gained a substantial level of expertise in this area.
iPaaS will step in as soon as the complexity has grown
For the initial integration requirements, while project implementers still take baby steps in IoT implementation projects, the embedded integration capabilities of IoT platforms are just enough to try out first integration scenarios and achieve first success. But as soon as IoT integration projects grow in complexity and involve integration with multiple applications and databases as well as more complex API management, this is when project implementers should introduce an iPaaS.
At the same time, Gartner points out that most likely, both IoT platforms with built-in integration capabilities AND stand-alone integration platforms will be used in the same IoT projects.
This actually fits with the observation I mentioned in my article “Ways to Manage Multiple Applications You Might Have Not Known About”: There are many integration platforms out there, but none of them represents a “silver bullet solution” for all integration needs. So, it is recommended to deploy several integration platforms that ideally complement each other.
I see IoT platforms / iPaaS being in the same situation. The technology is not mature enough to allow to build a solution that meets all IoT implementation & integration needs, and frankly speaking, this is also unnecessary. IoT platforms serve their own specific purpose, and their providers should work on honing the skills they already have. iPaaS providers have another, no less important set of expertise. A rational combination of several complementing expert tools is, therefore, a logical step to achieving the best results possible.
4 points to consider when choosing the right solution for your IoT integration projects1As already mentioned in the previous article, most IoT solutions provide APIs for a better interaction between them and other systems and applications. Gartner, however, stresses the importance of adopting the API-first, but by no means the API-only approach, pointing out that the latter is insufficient to address all integration requirements. (You’ll find a list of basic functional capabilities required for this in the research note covered in this article.) It makes sense to implement an iPaaS that pursues the same API-first approach to integration platform management. 2The IoT needs an event-driven architecture in order to ensure, among other things, an “efficient real time response and processing”. In fact, the event-driven architecture is increasingly utilized in IoT projects. Project implementers should, therefore, look for an iPaaS that supports this event-driven architecture. 3As with any other new technology, it is impossible to predict from the start what will eventual become out of an IoT implementation & integration project. It might start small and stop there, or it might grow big and extremely complex, involving not only SaaS and mobile solutions but the enterprise legacy systems. It is, therefore, important to include into your IoT strategy the possibility of not just adding some iPaaS at a certain step of the project, but adding the iPaaS that supports both SaaS integrations and on-premise integrations. If this is not an option, you might want to consider implementing a classic cloud-based iPaaS together with an on-premise integration middleware. 4Another point that is not mentioned in the research note, but still essential for IoT integration projects is the fact that an iPaaS should be able to provide very high scalability with low latency. This is crucial in order to be able to deal with parallel processing. Such performance can be achieved through the implementation of the microservices architecture. In other words, when selecting a stand-alone integration platform, IoT project implementers should first look for a microservices-based iPaaS.
I’d like to finish this article with a Gartner’s prediction regarding the usage of an iPaaS in IoT projects: By 2020 “75% of IoT projects will use some form of a stand-alone integration platform”.
I guess it means that it’s high time to start looking for a suitable iPaaS now, even if your IoT strategy is still in its infancy.