IoT devices collect an enormous range of data, but it falls into a few core categories: environmental readings (temperature, humidity, air quality), physiological and biometric data (heart rate, sleep patterns, blood pressure), behavioral data (movement patterns, appliance usage, driving habits), location data, and device metadata like IP addresses and timestamps. The specific data points depend on the device, but the underlying principle is the same: physical sensors detect changes in the real world and convert them into small, structured digital data points that get transmitted, stored, and analyzed.
Environmental Data
The most fundamental IoT sensors measure the physical world around them. Temperature and humidity sensors are everywhere, from smart thermostats in your living room to industrial monitors on a factory floor. Air quality sensors track concentrations of gases like CO2, detect UV radiation levels, and measure particulate matter. Pressure sensors monitor atmospheric conditions or, in industrial settings, the pressure inside pipes and hydraulic systems.
These readings are typically continuous. The sensor checks its environment at a set interval, sometimes hundreds of times per second, and logs the value each time. A smart home thermostat might record temperature every few minutes. An industrial sensor monitoring steam pressure on a boiler could sample dozens of times per second. Each reading is a small data tuple: a value, a timestamp, and a device identifier.
Smart Home Data
Inside a connected home, IoT devices build a surprisingly detailed picture of daily life. Motion sensors (passive infrared) track which rooms you occupy and when. Reed switches on doors and cabinets log every time you open the fridge or a kitchen drawer. Force sensors embedded in furniture can detect whether someone is sitting on the couch or lying in bed. Smart plugs record exactly when you turn on the coffee maker, dishwasher, washing machine, or microwave, and how much electricity each appliance draws.
Light sensors measure ambient brightness in different rooms. Smart speakers collect voice recordings and the wake words that triggered them. Security cameras capture video and, increasingly, use on-device processing to identify faces or detect specific objects like packages. Taken individually, each data point is mundane. Combined over weeks and months, they reveal detailed routines: when you wake up, when you leave for work, how often you cook, and how long you sleep.
Biometric and Health Data
Wearable IoT devices focus on the body. Fitness trackers and smartwatches commonly collect heart rate, step count, distance traveled, calories burned, and sleep duration and quality. More advanced medical wearables measure blood oxygen levels, blood pressure, skin temperature, and skin conductance (a proxy for stress). Clinical-grade IoT sensors go further, recording electrical activity from muscles, the heart, and the brain.
These devices generate continuous physiological streams. A heart rate sensor on a smartwatch might log a reading every second throughout the day. Sleep tracking combines motion data, heart rate variability, and sometimes blood oxygen to estimate sleep stages. The result is a longitudinal health profile that can reveal trends invisible to the wearer, like a gradual increase in resting heart rate or a slow decline in sleep quality over months.
Vehicle and Telematics Data
Connected cars are among the most data-rich IoT devices in everyday life. Onboard sensors and GPS units collect fuel level, battery status, tire pressure, engine diagnostics, and real-time location coordinates. Driving behavior data includes speed, acceleration and braking patterns, and how sharply you take turns.
In the event of a crash, the vehicle can transmit the speed at impact, the car’s position after the collision, and the number of passengers detected by seat sensors. Some systems also relay pre-crash health data from the driver, such as signs of drowsiness. Fleet management systems extend this further, tracking freight condition (temperature inside a refrigerated trailer, for example), alarm status, and maintenance schedules across hundreds of vehicles at once.
Smart City Infrastructure Data
At a city scale, IoT sensors feed into traffic management, waste collection, and environmental monitoring systems. Traffic sensors, for instance, log timestamps, street location, GPS coordinates for start and end points, vehicle counts, average speed, and measurement intervals. This data flows in real time, allowing traffic systems to adjust signal timing or reroute vehicles.
Air quality stations measure pollutant concentrations across neighborhoods. Smart waste bins report how full they are so collection trucks can optimize routes. Water systems monitor flow rates and pressure to detect leaks. Noise sensors map sound pollution. Each of these generates continuous, location-tagged data streams that city planners use for both immediate operations and long-term infrastructure decisions.
Industrial Machine Data
In factories and plants, IoT sensors monitor equipment health and production quality. Vibration sensors detect when bearings, shafts, or other rotating parts become misaligned or worn, flagging the need for maintenance before a breakdown occurs. Temperature and humidity sensors ensure products are manufactured or stored within required ranges. Pressure gauges are critical in industries that rely on gas, steam, or hydraulic systems.
Motion sensors measure spindle speeds, conveyor flow rates, and the movements of robotic arms. Force (weight) sensors control how much pressure automated arms or dispensing mechanisms apply. Even high-frequency electrical data from motors gets captured and analyzed for anomalies. Industrial IoT generates some of the highest data volumes of any sector because machines run continuously and sensors often sample at very high rates.
Personal Identity and Behavioral Data
Beyond direct sensor readings, IoT devices collect, and sometimes inadvertently expose, personally identifiable information. This includes obvious identifiers like your name, email address, and home address, but also more sensitive data: fingerprints and facial scans from smart locks and phones, voice prints from smart speakers, and precise geolocation trails from nearly any connected device with GPS or Wi-Fi.
Researchers categorize this personal data into four types. There is what you know (name, address, security questions), what you have (credit card numbers, account credentials), what you are (biometrics like fingerprints, facial geometry, blood type), and what you do (geolocation patterns, shopping habits, website visits). That last category is particularly revealing because IoT devices passively track behavioral patterns over time, building a profile of your routines without requiring you to actively input anything.
Metadata and Device Identifiers
Every IoT device also generates metadata: data about the data. This includes the device’s IP address, MAC address (a unique hardware identifier), browser fingerprints if the device has a web interface, and timestamps for every interaction. These identifiers allow networks to distinguish one device from another, but they also create a trail that can be used to track activity across systems.
Metadata might seem harmless compared to a heart rate reading or a video recording, but it carries real privacy weight. A log of timestamps and IP addresses from a smart lock tells anyone with access exactly when you come and go. Location headers from a fitness tracker reveal not just where you ran, but the route, the time, and how long you stayed at each point. Even aggregated, anonymized metadata can sometimes be reverse-engineered to identify specific individuals based on unique patterns of behavior.
How Raw Signals Become Data
At the hardware level, IoT sensors work by detecting a physical change and converting it into an electrical signal. A temperature sensor produces a voltage that varies with heat. A motion sensor outputs a simple on/off signal when it detects infrared radiation from a warm body. A vibration sensor generates a waveform that mirrors the mechanical oscillation of a machine part.
These analog signals get converted into digital values, typically represented as numbers, Boolean true/false states, or continuous data streams. The digital output is then packaged with a timestamp and device ID and transmitted to a gateway or cloud server. Some sensors do light processing at the edge, filtering noise or compressing data before sending it. Others stream raw readings for processing elsewhere. The format is almost always compact, because IoT networks are designed to handle millions of small transmissions rather than a few large ones.