Device integration: IoT’s Mount Everest
A steep challenge lies ahead for IoT
Integration testing is hard. Like climbing a mountain hard. If you’re a software developer – you already know this. IoT integration testing is even harder. If you’re an IoT software developer – you already live this.
The connectivity between different IoT devices is one area which remains a major challenge for middleware developers. While the value in having multiple synchronized IoT devices is obvious, integrating them so that they can seamlessly share data is another issue entirely.
According to Gartner, integration is among the top challenges facing IoT. “Through 2018, half the cost of implementing IoT solutions will be spent integrating various IoT components with each other and back-end systems.”
As the founder of a startup solving the problem of IoT integration, I was very interested in knowing if this is something other IoT developers are facing and, if so, how much pain it is causing them.
Zeinab Farahmandpour, a colleague of mine, recently attended the ICSE conference in Argentina to present our co-authored paper. In discussing the issue with attendees, she confirmed the commonality of this problem for IoT developers. A memorable quote she took away, “Setting up and maintaining such an interconnected environment for the development phase seems notoriously difficult, time consuming and very expensive.”
So, if this issue is notoriously difficult, how do we approach the climb? IoT device integration currently has only a few less-than-ideal options: inaccurate simulators, untrustworthy emulators or massively expensive real device labs.
KnowThings.io, a CA Accelerator incubation, has an entirely different approach: we attack this problem the same way climbers scale Mount Everest. By teaming up.
As an architect and product owner of CA Service Virtualization, my expertise lies in virtualizing fast-changing, dynamic environments. However, this is only one piece of the puzzle.
Even the most advanced climbers cannot expect to reach Everest’s peak alone. They require a sherpa – an individual with deep knowledge of the terrain who can guide them to the top. Virtualizing an IoT environment requires expertise across countless protocols – more than any developer can bring to the table. Therefore, we’ve turned to machine learning as our sherpa.
Thus, I’m excited that KnowThings.io is partnering with the machine learning experts at the Swinburne University of Technology in Australia. Together, we were the first to bring machine learning to CA Service Virtualization. We are applying the same approach to IoT, in order to create Adaptive Virtual Devices.
This solution will provide developers the power to virtually integrate IoT devices, allowing them to shift-left IoT development.
IoT integration testing and software development present an enormous challenge. KnowThings brings the experience and learning-based technologies to navigate this journey.
The summit awaits.
If you’d like to learn more about what this solution can do for you, visit KnowThings.io.