Robots Atlas>ROBOTS ATLAS
Infrastructure

IoT

1999ActivePublished: 21 October 2026Updated: 21 October 2026Published
Key innovation
Giving physical objects a network identity and the ability to exchange data, turning the material world into a real-time source and consumer of data without a human in the loop.
Category
Infrastructure
Abstraction level
Paradigm
Operation level
DataDeploymentInference
Use cases
Telemetry and remote monitoring of robot fleetsPredictive maintenance of machineryIndustrial IoT (IIoT) and automation within Industry 4.0Smart home and smart buildingsSmart cities โ€” lighting, parking, environmental sensingEdge AI / AIoT โ€” running model inference on edge devicesPrecision agriculture and environmental monitoringLogistics and asset tracking

How it works

1) Data capture: sensors on devices measure physical quantities (temperature, motion, position, energy use, imagery) and convert them into digital data. 2) Connectivity: data is transmitted over a network โ€” wirelessly (Wi-Fi, Bluetooth, cellular, LoRaWAN) or wired โ€” often using lightweight messaging protocols such as MQTT or CoAP, optimized for constrained bandwidth and power. 3) Processing: data reaches a platform (cloud or edge server) where it is aggregated, analyzed, and stored; part of the processing and AI inference may run locally on the device or gateway (edge computing) to cut latency and reduce transfer. 4) Action and feedback: based on the analysis the system makes decisions โ€” presenting them to a human, triggering automated rules, or sending commands to actuators, closing the control loop between the digital and physical worlds.

Problem solved

It solves the isolation of the physical world from digital systems. Before devices gained connectivity, data about the state of machines, environments, or products had to be collected manually and with delay. IoT enables continuous, automatic measurement, reporting, and remote actuation of physical objects, forming the basis for real-time monitoring, predictive maintenance, and data-driven control.

Components

Device layer (sensors and actuators)Acquiring data and acting on the physical world

The physical endpoints of an IoT network: sensors that measure the environment (temperature, motion, imagery, position) and actuators that act on the world (motors, valves, relays). This is where physical data becomes digital.

Official

Connectivity layerTransporting data across the network

The networks and protocols moving data between devices and the platform: Wi-Fi, Bluetooth, cellular (LTE/5G), LoRaWAN, and lightweight messaging protocols such as MQTT and CoAP, optimized for constrained bandwidth and power.

Official

Platform / application layerAggregating, analyzing, and serving data

The cloud or edge servers where data is aggregated, stored, analyzed, and exposed to users and other systems. This is where automation rules, dashboards, and AI models processing device streams run.

Official

Evolution

1999
Kevin Ashton coins the term "Internet of Things"
Inflection point

During a Procter & Gamble presentation on using RFID in the supply chain, Ashton used the phrase "Internet of Things".

2008
Connected devices outnumber people

According to Cisco IBSG estimates, around 2008-2009 the number of internet-connected devices surpassed the human population โ€” a symbolic "birth" of IoT as a mass phenomenon.

2017
Rise of AIoT and edge AI

In the second half of the 2010s, combining IoT with AI inference at the edge (AIoT) became widespread, moving part of model processing directly onto devices and gateways.

Hardware requirements

Primary

IoT is a hardware-agnostic paradigm โ€” it spans ultra-low-power microcontrollers, edge gateways, and cloud servers. Hardware choice depends on the application, not on the concept itself.