UK What parking occupancy sensors do
Parking occupancy sensors detect whether a parking space is free or taken and report that status to a central system. Each sensor watches one bay, or in the case of a camera, a group of bays. Together they give an operator a live map of the car park.
On their own the sensors only answer one question: is there a vehicle in this space? The value comes from what is done with the answer, such as guiding drivers to a free bay, enforcing time limits, or planning capacity from months of occupancy data.
How each type detects a vehicle
Different sensors detect a car in different ways, which is why they suit different sites.
-
//
Ultrasonic. Sends a pulse of sound down to the floor and times the echo. A car roof returns the echo sooner than the floor does.
-
//
Magnetometer. Measures the change a car’s steel body makes in the Earth’s magnetic field above the sensor.
-
//
Radar. Sends a low-power radio signal upwards and detects the reflection from the underside of a vehicle. Often combined with a magnetometer to cut false readings.
-
//
Infrared. Detects the presence of an object by reflected or interrupted infrared light.
-
//
Camera with AI. A vision model classifies each marked bay in the image as occupied or free, frame by frame.
Parking sensor types compared
The table compares the common sensor types. Real installations often mix them, for example ultrasonic sensors on covered levels and cameras on the open roof deck.
| Sensor type | Mounting | Connection | Best for | Watch out for |
|---|---|---|---|---|
| Ultrasonic | Ceiling, above each bay | Wired bus to a zone controller | Indoor multi-storey car parks | Needs a ceiling and cabling to every bay |
| Magnetometer | Surface or core-drilled into the bay | Battery, wireless (LoRaWAN, NB-IoT) | Outdoor and on-street bays | Interference from nearby vehicles and metal |
| Radar | In-ground or overhead | Battery wireless or wired | Outdoor bays, often paired with a magnetometer | Higher cost per bay |
| Infrared | Overhead or wall | Wired | Short-range indoor detection | Affected by dirt and strong light |
| AI camera | Pole, wall or ceiling | PoE to an edge computer | Surface car parks and large decks | Needs a clear view and processing |
Choosing sensors for the site
The structure of the car park usually decides the sensor. These are the usual starting points:
-
//
Indoor multi-storey. Ultrasonic sensors with bay lights are a proven option.
-
//
Outdoor surface car parks. In-ground magnetic or radar sensors avoid poles and cabling.
-
//
Large open areas. Cameras cover many bays and add video evidence.
-
//
On-street parking. Battery-powered in-ground sensors with a low-power wireless network, because there is nowhere to run cables.
-
//
Processing. Plan an edge computer to collect sensor data or run camera analytics.
Camera-based occupancy detection
Camera-based detection has become popular because one camera can watch many spaces and the same images can be used for ANPR and security. It needs more processing, usually an edge computer running a vision model, while ultrasonic and magnetic sensors send a simple occupied or free signal.
A typical design powers the cameras over Ethernet and connects them to a PoE computer that runs the occupancy model locally and sends only the bay status upstream. Our video analytics FAQ explains the AI side, and our guide to PoE surveillance and NVR systems covers power budgets and recording. Our smart parking case study shows PoE systems running a live car park.
Installing and maintaining sensors
Installation cost and upkeep often matter more than the price of the sensor itself, so plan them from the start.
-
//
In-ground sensors. Fitted in a core-drilled hole or bonded to the surface. Plan the work around resurfacing and line painting, and check how the battery is replaced.
-
//
Overhead sensors. Need a ceiling at a suitable height and a cable run to every bay, so they are easiest to fit during construction or refurbishment.
-
//
Cameras. Need a clear, stable view, cleaning of the housing and a site survey for lighting at night.
-
//
Calibration. Most systems are tuned after installation, then rechecked after layout changes or when false readings appear.
-
//
Network coverage. Wireless sensors need a survey to confirm every bay reaches a gateway, including those under parked lorries or ramps.
Where the occupancy data goes
Occupancy data is collected by zone controllers, gateways or edge computers and passed to the parking software. From there it drives count signs and bay lights, which our parking guidance FAQ describes, and it feeds the wider car park management system that handles access, payment and reporting. Many operators also publish live availability to apps and city information services.
Quick answers on parking occupancy sensors
Parking occupancy sensors: common questions
How accurate are parking occupancy sensors?
High when installed and tuned well. Accuracy depends on the sensor, the site and the calibration. Trial a few bays before a full roll-out, especially near steel structures or busy roads.
How long do in-ground sensor batteries last?
Often several years. Battery life depends on how often the sensor reports and how busy the bay is. Check the manufacturer’s figure for your expected turnover.
How many bays can one camera cover?
Anything from a few to dozens. It depends on the mounting height, lens and angle. Vehicles must not hide each other from the camera’s view.
How do wireless parking sensors connect?
Usually over a low-power network. LoRaWAN, NB-IoT and similar networks carry small status messages from battery-powered sensors to a gateway.
Is counting the same as occupancy sensing?
No. Counting tracks vehicles in and out of a level or zone. Occupancy sensing checks each bay, so it can guide drivers to a specific space.
We supply PoE and edge AI computers for parking projects with the camera inputs, network and operating system configured as standard. Tell us the number of bays, the sensors or cameras and the software, and we will specify the computing to match.