Stanley says that in some circles, by definition “M2M is generally a many to one situation” where “the device has an affinity to one host” and does not connect to multiple places or make intelligent decisions about where it will connect. An example of this would be automated meter reading systems in increasing use in the electric utility vertical. The meters, in this case, aren’t making intelligent decisions about where to send data; they just monitor and transmit to a fixed point. This type of application can exist in any number of vertical industries where devices are collecting and transmitting data on anything from logistics and telemetry to security and health or network monitoring.
But limiting M2M’s definition to purpose-built networks of devices that don’t make intelligent decisions, or don’t have the ability to make autonomous decisions about where to send certain types of data, would seem to overlook both reality and opportunity. What we’re seeing develop in consumer apps provides a glimpse into what’s possible. In a sense, the apps on one’s smartphone or tablet could themselves be considered devices or machines; they may be software-based, but they’re still machines.
We already see examples of apps that can operate autonomously to back up data, photos and videos from a smartphone to the cloud. Once the human user configures the settings, the device can act on its own. It’s not a stretch to think that a user could tell the device to post videos to YouTube, photos to Facebook, and emailed documents to DropBox.
The human may not push another button or do anything other than shoot photos and video during a week-long family trip to Orlando; but every night her smartphone uploads her files on its own.