Connected Sensor: What is it?

What is a Connected Sensor?

A connected sensor (or smart sensor) is a device that combines a physical measurement element (transducer), an embedded processing unit (microcontroller), and a network communication module to automatically collect, pre-process, and transmit environmental data to a centralised system. Unlike a traditional sensor that requires manual reading, a connected sensor sends its measurements continuously or at programmed intervals, enabling permanent monitoring without human intervention.

Connected sensors measure a wide variety of physical quantities: temperature (rooms, equipment, outdoor), relative and absolute humidity, atmospheric pressure, air quality (CO2, fine particles PM2.5/PM10, VOCs), light levels, motion and presence (PIR, radar), open/close status (magnetic contact), vibration (accelerometer), level (liquids, tanks), energy consumption (current, power), and position (GPS, triangulation).

Communication is carried out via protocols suited to the use case: Zigbee or Bluetooth for short indoor distances, LoRaWAN for long outdoor distances, Wi-Fi for environments with existing network infrastructure, and NB-IoT/LTE-M for cellular deployments. Data then transits via MQTT or HTTP to the application backend (Django) for storage, analysis, and visualisation.

Why Connected Sensors matter

Connected sensors are the foundation of any IoT solution. Without them, there is no field data, no monitoring, no automation. Their strategic importance manifests across multiple domains.

• Operational visibility: sensors transform invisible physical phenomena (air quality, vibrations, energy consumption) into measurable, actionable data, eliminating management blind spots.
• Predictive maintenance: by continuously monitoring equipment parameters (temperature, vibration, current), sensors detect early warning signs of failure and enable intervention before breakdown.
• Energy efficiency: temperature, presence, and light sensors in buildings allow heating, air conditioning, and lighting to be adapted to actual occupancy, reducing consumption by 20 to 40%.
• Regulatory compliance: in the food, pharmaceutical, or hospital sectors, connected sensors ensure continuous, timestamped recording of environmental conditions, facilitating audits and compliance with standards (HACCP, GMP).
• Data-driven decision making: sensor data feeds dashboards and analyses that transform intuition into informed, measurable, and reproducible decisions.

How it works

The operation of a connected sensor follows a four-step cycle. Measurement begins with the transducer converting a physical quantity into an electrical signal. A resistive temperature sensor (NTC) changes its resistance based on temperature, a PIR motion sensor detects infrared radiation from warm bodies, and a particulate matter sensor uses laser diffraction to count suspended particles.

The electrical signal is then digitised by the microcontroller's analogue-to-digital converter (ADC). For modern digital sensors (BME280, SCD40, SEN55), digitisation and calibration are integrated within the sensor itself, which communicates directly via I2C or SPI with the microcontroller. The microcontroller (Arduino, ESP32) or single-board computer (Raspberry Pi) applies pre-processing: calibration, noise filtering, unit conversion, and data formatting.

Transmission sends the formatted data (typically in JSON or CayenneLPP for LoRaWAN) via the appropriate communication module. Protocol choice depends on distance (Zigbee: 10-100 m, LoRaWAN: 2-15 km), transmission frequency (Zigbee: seconds, LoRaWAN: minutes), power consumption, and data volume. The backend (Django) receives data via an MQTT broker, stores it in PostgreSQL, and exposes it via REST APIs for dashboards and alerts.

Concrete example

At Kern-IT, we deploy connected sensors across many sectors for our Belgian clients. For a smart building project, we installed a sensor network comprising temperature/humidity probes (Zigbee) in every office, CO2 sensors (LoRaWAN) in meeting rooms, presence detectors (Zigbee) for automatic lighting, and connected energy meters on electrical panels.

The entire sensor network converges on Raspberry Pi gateways (one per floor) that publish data on MQTT to our Django backend. Data is stored in PostgreSQL with TimescaleDB and visualised on KERN MAP with a floor-by-floor cartographic building view. The manager accesses a real-time dashboard showing occupancy, air quality, and energy consumption, with automatic alerts when CO2 exceeds 1,000 ppm (ventilation recommendation) or temperature falls outside the comfort range (19-24 C).

Implementation

1. Identify quantities to measure: precisely define which physical parameters are relevant to your use case and the required measurement ranges, precision, and sampling frequency.
2. Choose sensors: select sensors suited to the environment (IP67 for outdoor, ATEX-certified for explosive environments) and communication protocol (Zigbee for dense indoor, LoRaWAN for long-range outdoor).
3. Plan power supply: determine whether sensors will be mains-powered (simple maintenance but wiring required), battery-powered (autonomy to optimise), or energy-harvesting (solar, vibration).
4. Deploy gateways: install Raspberry Pi devices with appropriate coordinators (Zigbee dongle, LoRaWAN concentrator) to cover the deployment area.
5. Configure the data chain: set up the sensor → gateway → MQTT → Django → PostgreSQL → dashboard flow, testing each step individually.
6. Calibrate and validate: compare sensor measurements with reference instruments, adjust calibration offsets, and validate accuracy over a representative period.

Associated technologies and tools

• Zigbee / LoRaWAN: wireless communication protocols for sensor networks, suited to indoor and outdoor deployments respectively.
• Raspberry Pi: collection gateway that aggregates data from multiple sensors and transmits it to the central backend.
• MQTT: lightweight messaging protocol for reliable sensor data transmission to the application backend.
• Django / DRF: backend framework for storing, processing, and exposing sensor data via REST APIs.
• PostgreSQL / TimescaleDB: relational database with time-series extension for optimised sensor time-series storage.
• KERN MAP: Kern-IT's interactive mapping platform for geospatial visualisation of sensor data.

Conclusion

Connected sensors are the first building block of any IoT solution, transforming the physical world into actionable digital data. Choosing the right sensor, communication protocol, and data architecture determines the success of any monitoring project. At Kern-IT, we master the entire chain, from field sensor to dashboard, combining Zigbee and LoRaWAN for connectivity, Raspberry Pi for the gateway, MQTT for transport, Django for the backend, and KERN MAP for visualisation, delivering turnkey, reliable, and scalable monitoring solutions to our Belgian clients.