Bernd Gross believes the IoT industry has reached a turning point where the focus is now shifting to deployment. It's hard to argue with, around the world there's notably less talk and some impressive use of IoT technologies to solve real problems.
“After many years of piloting, proof of concepts… we’re now seeing real commercial products,” says Gross. “We’ve experienced that from our side at least twelve months ago.”
SoftwareAG has many high profile partners including the likes of Deloitte, IBM, Huawei, and many more. With such vast experience, Gross and his colleagues believe they have identified patterns to identify which projects are most likely to succeed.
Gross recommends a steady approach to adopting the IoT into a project, with three key stages – Data-driven actions, data-driven process integration, and, finally, learn and innovate.
“Run applications in a discrete, separate fashion,” advises Gross. “Do not integrate them before you have stabilised your new world of IoT. You have a different point of failure – if you change processes, you may end up falling into the category of failed IoT projects. Once you have it under control – six months, nine months … you can start changing your existing processes.”
Many projects can jump into machine learning and AI too soon. It's understandable, unlocking things like predictions can provide a major business advantage – but it’s ultimately worthless if you’re setting yourself up to fail.
Instead, Gross suggests ensuring your project moves steadily through the stages to ensure it’s stable. Adding things like machine learning could be a year or so down the line, but it will be worth it to have a solid base from which to innovate.
To build that solid base, it’s important to make three key considerations:
The first is the ability to deploy a minimum viable product quickly as in today’s market it’s expected in weeks, not months. Next is to ensure platforms are open and flexible to ensure you avoid ‘vendor lock-in’ and can deploy to any environment you want – whether that’s AWS, Google Cloud, Azure, Alibaba, or any other. Finally, it must be secure.
Security is of paramount importance and it goes beyond using a safe platform, the actual ‘thing’ must be secure. The Mirai botnet has become the go-to example of what happens when things are not secured – but if not taken seriously, it won’t be the last.
Gross’ suggestions will help any company looking to get into the IoT not to become one of the 65 percent which fails. Listen to his advice, and use it to avoid becoming another statistic.
What are your thoughts on Gross’ advice? Let us know in the comments.