oneM2M: One Standard for Connected Machines

Published May 23, 2026

The early Internet of Things had a fragmentation problem. Every vertical built its own stack, so a smart meter platform had nothing in common with a fleet tracker, and each new application meant another round of custom integration. oneM2M was set up to deal with that. The original workshop showcased companies using the standard; this page explains what the standard is.

Who is behind it

oneM2M was launched in 2012 by a group of regional standards bodies from Asia, Europe and North America, joined by industry forums and companies from many sectors. Its first full release appeared in early 2015, which is why it featured so prominently at conferences that year. Later releases have added features and tightened alignment with cellular IoT.

What it specifies

oneM2M does not define a radio or a network. It defines a common service layer that sits between applications and whatever connectivity happens to be underneath, whether that is cellular, Wi-Fi, LoRa or wired Ethernet. The layer provides functions that every IoT project ends up needing:

• Device registration and identification
• Data storage and sharing
• Subscriptions and notifications
• Access control and security
• Device management

These capabilities live in a Common Services Entity, which can run in the cloud, on a gateway, or on a capable device. Applications reach it through a RESTful interface and work with a tree of resources. Protocol bindings map those operations onto HTTP, CoAP and MQTT, so that a tiny sensor and a large server can take part in the same system.

Why a shared layer helps

Imagine a city with parking sensors, streetlights and waste bins from different suppliers. Without a common layer, each comes with its own platform and its own dashboard. With one, a single application can subscribe to data from all three. Vendors can also compete on devices and services without trapping the customer inside one silo.

Machine-to-machine before the name

Some industries built machine-to-machine systems long before the term was fashionable. Alarm panels phoned home over copper lines, and utilities read meters by radio. Gaming offers a neat example: slot machines on a casino floor have reported their meters to central accounting systems for decades, using protocols such as SAS. That kind of closed, single-purpose network is exactly what the oneM2M founders wanted to move beyond, although a regulated industry has good reasons to be cautious about opening one up.

Where things stand

oneM2M has found its strongest uptake in government-backed smart city projects, particularly in South Korea and India, and in some utility and industrial settings. It has not become the single layer for all IoT, as cloud providers' own platforms and other open protocols took much of the market. Even so, the case it made has aged well: the cost of an IoT project is rarely the sensor. It is the integration around it.