Our partner, Libelium, recently conducted a survey into the top 10 challenges of deploying IoT. It was well-sized with 600+ participants across a broad set of users and verticals, and the typical participant was an SME/SMB company doing an IoT project of 6m - 1y duration, involving 304 people and costing less than €1million.
Here are our favourite learnings:
Cost-saving is a key motivation
Interestingly, the principal motivation for projects was usually cost-saving. The survey is silent as to why, but from our own experience at DevicePilot, it's likely this is mixture of internal cost-saving (delivering a product or service more efficiently) and external cost-saving (enabling customers to deliver their offerings more efficiently). And 81% said that data can create a new revenue source. So there's a clear "increase revenue, lower costs → greater profit" message.
There are three common main challenges
The biggest three challenges reported were connectivity, hardware integration, and cloud integration, with the former being dominated primarily by the fixed characteristics of local and national geography (i.e. what connectivity is actually possible).
We continue to move into an 'IoT without software' ecosystem
73% of respondents wanted sensor nodes to be pre-programmed, which seems to fit well with the idea that we're moving into a world where it's possible and desirable to deploy IoT without having to write any software.
Standards-based connectivity is winning
There's a complete smorgasbord of connectivity types, though it's notable that 87%+ of companies are now using standards-based connectivity rather than rolling their own, which seems like a big step forward.
Your first IoT deployment is always the hardest
68% of respondents agreed with the statement that "it's difficult to define the parameters to be measured in order to achieve the project goals"... and to select sensors to achieve that. This chimes with our view that it's usually impossible to be completely right with your first IoT deployment because there are too many unknowns - you can either get stuck in paralysis by analysis, or you just strike out with your project and accept that it'll be a learning process.
Do we need better professionals or better tools?
Finally, it's surely significant that 98% of respondents say that the market needs professionals with a higher level of knowledge. From DevicePilot's perspective, that reinforces the idea that it's an advancement if your tools can help free you from the need for that knowledge.