I am rather surprised you mixed Google and AWS with Adafruit.io, which in my opinion, is the only IoT Service listed. I would consider Google and AWS to be more akin to IoT platforms.
You forgot to add the catch-all "Other" category. Also what about those who do use these services. There is no tick box for that category of users.
One aspect that might put people off would simply be "complexity". That is certainly the case with AWS for the majority of simple data capture applications. In those use cases, I found that people usually gravitate towards something like Thingspeak.
This isn't a 'what puts you off' response, but more of a 'gotcha', is that some services are not mature, and can take more messing around than desired.
Right now, nothing is putting me off it, because there are problems I cannot solve without it. I want to develop a monitoring solution for air conditioners (HVAC) in business premises - there's no way I intend to run my own server. It doesn't make sense, that problem (and security) was solved with cloud computing 10 years ago - it's why we all use it today whether we like it or not, each time we connect to any service to get work done.
It is the age old tug-of-war between big business that would like to centralize everything and maintain control of a continuous revenue stream; and the individual who wants to have local control and no service fee obligations. Big business invests in centralized services that are attractive enough for end users to subscribe, but eventually the cost comes down to the point where end users can get what they want without subscribing. Not everyone drops out of centralized service models, but there is always demand for local control.
Mainframe computers with end user terminals were king until PCs got enough power to eliminate the need for mainframes.
Then servers with thin clients had a run until memory and disk prices came down to make servers less needed.
Now cloud computing is having a go while central computing can provide better access and analysis of big data.
Presumably phones will eventually have enough power to do much of it locally.
An example is voice recognition. Right now cloud computing has better voice recognition, but off-line voice recognition is rapidly catching up.
Do users want their phone to perform voice recognition or have to connect to the cloud to do it?
I expect most end users will want local control unless they see a compelling application provided by the cloud, but if they could get the same performance locally they would prefer it.
Business decisions not to roll-their-own obviously tend to be made by managers who cannot "roll their own".
Probably some of 1 and 2 from your list, I would worry I would likely forget I had subscribed and run up a large bill.
However for me it would more likely be security related. Not because I know those services are unsafe to use but because I don't understand them fully and are naturally suspicious. I'd be mortified if having taken part someone pointed out that my entire account was insecure or that I'd managed to post some personal data/leave the account open etc. E.g. I log out of every application when I finish using it....that's me !
It's uncertainty over the cost rather than the cost itself. When using these services, I always delete the resource used as soon as I have finished testing as the cost seems to mount up inexorably. It always seems to cost more than I thought it would before I signed up for it.
Top Comments