The unifying power of natural language

About 70,000 years ago, in order to satisfy their need to communicate, Sapiens developed what is now called "natural language". In the same way, natural language is now a good support for collaboration within the IoT
IoT is a universe made of silos not ready to interoperate

In IoT, interoperability usually means providing an API. However, this is not the end of the story as each individual component has its own API. Each member of the group must then learn the language of all the others, which makes collaboration clumsy and neither flexible nor scalable.


The IoT must evolve from the individual and centralized connection of devices to an open collaboration model supporting a growing diversity of workflows, devices, platforms, applications and users.

Interconnecting APIs in a one-to-one manner is not scalable (complexity evolves in n2)

Developing and maintaining a collection of APIs requires time from scarce technical resources

Poor interoperability is at the origin of an ever growing technical debt


Place of interactions for a collaborative IoT

Pragmatics (meaning in the context)

Semantics (literal meaning)

Syntax (phrases and sentences

Morphology (words)

Phonology (phonemes) Phonetics (sounds)

a common language

Collaboration between things, applications and people requires a unique language, quickly acceptable to all and compatible with the functioning of the connected objects of today and tomorrow. This language must be flexible to allow unsuspected future uses and stable enough to ensure the longest possible lifespan for the infrastructures that will implement it.


Standardization bodies (oneM2M, W3C) have defined semantics as the most convenient level of interoperability: diversity and continuous changes are in the nature of things, making standardization at lower layers intractable. Consequently, general interoperability must be built at a higher level: semantics.  

Natural language for IoT collaboration

Natural language allows interactions at the semantic level: its flexible syntax incorporates the meaning we want to convey. Its success is due to its versatility and the dissociation between semantics and syntax: there are many ways to express the same intention, which allows to focus on the meaning rather than on the strict structure of sentences.

Note that natural language is already present in IoT: connected objects and applications are specified and documented by their (human) designer in natural language, the transition to more structured (machine) languages being made necessary due to the current state of technology.

Semantics-level communication, common to machines, application and users

Already in use, collectively defined and belonging to all; no obsolescence in sight

Leverages existing communication services (sms, collaboration platforms, voice assistant, etc.).