IoT versus machine to machine

One common area of confusion in the IoT world is what separates it from the technologies that defined machine to machine (M2M). Before IoT became part of the mainstream vernacular, M2M was the hype. M2M and IoT are very similar technologies, but there is a significant difference:

• M2M: It is a general concept involving an autonomous device communicating directly to another autonomous device. Autonomous refers to the ability of the node to instantiate and communicate information with another node without human intervention. The form of communication is left open to the application. It may very well be the case that an M2M device uses no inherent services or topologies for communication. This leaves out typical internet appliances used regularly for cloud services and storage. An M2M system may communicate over non-IP based channels as well, such as a serial port or custom protocol.
• IoT: IoT systems may incorporate some M2M nodes (such as a Bluetooth mesh using non-IP communication), but aggregates data at an edge router or gateway. An edge appliance like a gateway or router serves as the entry point onto the internet. Alternatively, some sensors with more substantial computing power can push the internet networking layers onto the sensor itself. Regardless of where the internet on-ramp exists, the fact that it has a method of tying into the internet fabric is what defines IoT.  

By moving data onto the internet for sensors, edge processors, and smart devices, the legacy world of cloud services can be applied to the simplest of devices. Before cloud technology and mobile communication became mainstream and cost-effective, simple sensors and embedded computing devices in the field had no good means of communicating data globally in seconds, storing information for perpetuity, and analyzing data to find trends and patterns. As cloud technologies advanced, wireless communication systems became pervasive, new energy devices like lithium-ion became cost-effective, and machine learning models evolved to produce actionable value. This greatly improved the IoT value proposition. Without these technologies coming together when they did, we would still be in an M2M world.