Pass event and lane
Detection of each vehicle and the lane it travels in as it passes the measurement point.
Technical guide · Mobility & ITS
Vehicle counting and classification systems: how to choose the right technology
Compare radar, cameras, inductive loops, pneumatic tubes, Bluetooth/WiFi and manual counting by available data, installation, privacy, maintenance and integration.
A general guide for urban mobility, road, industrial-area and Smart City projects.
FMCW Radar
Cameras
Loops
Tubes
Manual counting
Quick summary: which technology fits best
Pneumatic tubes
For temporary survey campaigns , they can be enough.
Cameras
For visual identification or licence plates (LPR).
Inductive loops
For permanent in-road installations.
FMCW Radar
For continuous counting with no cameras and no civil works.
Concept
What is a vehicle counting system?
A vehicle counting and classification system is a set of sensors and software that detects the vehicles passing a point on the road, counts them and groups them by type. Depending on the technology, it can add information on speed, direction of travel, lane and occupancy..
This data characterises how a road behaves over time: analysing peak hours, calculating annual average daily traffic and providing objective information for urban planning and mobility management.
This guide explains the main vehicle counting and classification technologies and how to choose between them. For models, specifications and radar-based installation, see the uRAD Smart Traffic solution.
- Vehicle counting
- Classification by type
- Speed
- Direction & lane
Data flow
How a vehicle counting system works
Road userVehicle, bike or pedestrian
SensorRadar, camera, loop…
ProcessingDetection and classification
Structured dataLane, direction, speed
ITS / Smart City platformAnalysis and integration
Whatever the technology, the system turns pass events into structured data for analysis, planning or integration with external platforms.
Available data
What data can a counting system provide?
Depending on the technology and project configuration, a counting and classification system can provide different types of information.
Data captured directly
Information the sensor captures on each pass event.
Direction of travel
Distinguishes vehicles approaching from those moving away from the measurement point.
Speed
Speed of each vehicle or section average, depending on the technology used.
Classification by type
Grouping by type: pedestrian, bicycle, light or heavy vehicle.
Indicators calculated from the data
Metrics derived by aggregating and analysing the captured events.
Annual average daily traffic (AADT)
Aggregated traffic volume that helps size the road.
Hourly distribution and peak hours
Demand curves across the day and identification of traffic peaks.
Weekly trend
Comparison of volumes across days and weeks to spot trends.
Congestion indicators
Degree of road saturation to anticipate and manage congestion.
Available technologies
Technologies for counting and classifying vehicles
Each technology offers a different trade-off between the data it provides, privacy, installation and maintenance. This is a balanced overview.
FMCW Radar
Frequency-modulated continuous wave
Emits radio waves and measures the reflected signal to detect presence, distance, speed and direction without capturing personal images.
Best for24/7 counting, speed and direction with no civil works or cameras.
BenefitsDay/night, robust in rain or fog, low maintenance.
LimitsSensitive to geometry, occlusions and sensor position.
AI cameras
Video analytics
Process images to detect and classify vehicles, with licence-plate reading depending on the system.
Best forWhen visual detail or licence plates (LPR) are needed.
BenefitsDetailed classification and visual verification of events.
LimitsPerformance affected by lighting, optics and weather depending on the system; they involve image processing.
Inductive loops
In-road coils
Coils embedded in the pavement that detect a vehicle’s metallic mass as it passes.
Best forPermanent installations at defined fixed points.
BenefitsRequires civil works and replacement; traffic disruption.
LimitsRequires civil works and resurfacing; road closure.
Pneumatic tubes
Temporary counts
Tubes across the road that record axle passes via air pulses.
Best forTemporary studies and short-duration survey campaigns.
BenefitsFast deployment and low cost for one-off measurements.
LimitsWear and tear; ill-suited to dense or permanent traffic.
Manual counting
Operator counting
Staff who count and classify vehicles visually, on site or from video.
Best forVery specific studies or spot validation of data.
BenefitsHuman judgement for complex or unusual classifications.
LimitsNot continuous, labour cost and possible fatigue.
Bluetooth / WiFi
Signal detection
Detect on-board devices to estimate flows and travel times. It is a complementary technology, not direct counting and classification.
Best forTravel times and origin-destination matrices.
BenefitsUseful for network-level mobility studies.
LimitsComplementary technology: estimates flows, does not perform exhaustive counting or direct classification.
Comparison
Comparison table of vehicle counting technologies
A guide overview placing each technology by its best use, strengths, limitations, privacy and installation.
| Technology | Best use | Benefits | Limitations | Privacy | Installation |
|---|---|---|---|---|---|
| FMCW Radar | Continuous counting with speed and direction | Day/night, weather-robust, low maintenance | Sensitive to geometry and occlusions | No personal image | Non-invasive, no civil works |
| Cameras | Visual classification and license plate | Rich data, visual verification | Affected by lighting, optics and weather depending on the system | Captures images | Medium, requires calibration |
| Inductive loops | Permanent counting at a fixed point | Mature and reliable per lane | Civil works and resurfacing | No image capture | Invasive (road opened) |
| Pneumatic tubes | Temporary survey campaigns | Fast and low-cost | Wear; not for dense traffic | No image capture | Temporary on the road |
| Manual counting | One-off studies / validation | Flexible human judgement | Not continuous, staff cost | No automatic recording | No fixed installation |
| Bluetooth / WiFi | Travel times and O-D | Network-level view, easy to deploy | Complementary: does not count or classify exhaustively | Subject to minimisation and pseudonymisation | Simple, on existing infrastructure |
Guide table: the real performance of each technology depends on the manufacturer, model, configuration and conditions at the measurement point. Field validation of the data is recommended.
When to use
When does radar make sense?
Radar can be especially useful where continuity, privacy and non-invasive installation are priorities.
- Continuous 24/7 counting
- Cameras are not wanted
- Reducing personal image processing
- Measuring speed and direction
- Avoiding invasive civil works
- Day and night operation
- Smart City or ITS projects
- Integration with external platforms
Radar and camera
Radar and camera don’t solve exactly the same problem
Radar
- Detects presence, distance, speed and direction.
- Generates no personal imagery.
- Works day and night.
- Depends on geometry and sensor position.
vs
Camera
- Captures images for visual analytics.
- Can verify events or read licence plates.
- Depends on light, weather and image processing.
- Useful when imagery is required.
If the project needs visual identification or licence plates, a camera may be right. If the goal is counting, speed and direction without capturing images, radar is usually an option worth considering.
Privacy
Privacy and camera-free counting
In urban projects, image processing can be a relevant factor. Technologies that capture no images, such as radar, present a different privacy profile from video-based solutions.
- No capture of personal images of pedestrians, drivers or licence plates.
- Lower exposure to identifiable data compared with video systems.
- Can simplify some deployment scenarios in sensitive areas.
- Can reduce friction in urban settings with public sensitivity.
Compliance always depends on the applicable regulation, on how data is processed and on the specific configuration of each project. This page does not constitute legal advice.
Selection criteria
What to define before choosing a system
This checklist helps scope the project and prepare a feasibility enquiry or quote request.
Deployment: temporary or permanent
Number of lanes to cover
Directions of travel
Need for classification
Need for speed
Whether v civil works are allowed
Privacy requirements
Power supply
Connectivity on site
ITS or municipal integration
Environmental conditions
Accuracy accuracy and maintenance
Limits and installation factors
Limitations and factors affecting accuracy
No system is perfect in every scenario. Knowing these factors helps plan realistic deployments and interpret the data well.
Why installation shapes accuracy
Height & angle
Detection zone
Lanes & direction
Occlusions
Accuracy does not depend on the technology alone. Sensor position, road geometry, lane spacing and occlusions all affect the result.
Sensor position and height
Poor mounting affects the covered area and lane separation.
Road geometry
Curves, gradients and width shape detection and classification.
Very dense traffic
Congestion can make it harder to separate vehicles close together.
Occlusions
Large vehicles can hide smaller ones behind them.
Static objects
Street furniture or parked vehicles can add noise to the signal.
Mount vibration
Unstable masts or poles can affect measurement quality.
Extreme weather
Very adverse conditions can affect results to varying degrees by technology.
Calibration needed
Initial setup is key to ensuring the expected accuracy.
Field data validation
Results should be checked against reference measurements.
In practice, the real accuracy of any system requires field validation and depends on the configuration and the specific environment of the measurement point.
Use cases
Where vehicle counting systems are used
Different contexts have different data needs; the right system depends on the goal of each project.
Local authorities
Objective data to manage urban traffic and prioritise actions.
Smart Cities
Counting integrated into urban platforms for data-driven decisions.
Highways
Characterising volumes and speeds on interurban roads.
Industrial areas
Access control and heavy-vehicle monitoring in logistics.
Logistics access
Counting at entries and exits of high-turnover facilities.
Cycling infrastructure
Counting bicycles and vulnerable users to size and assess bike lanes.
Temporary studies
Mobility campaigns to measure the impact of an intervention.
Urban planning
Evidence to size road infrastructure.
Next step
From choosing a technology to a concrete solution
If, after comparing technologies, you need continuous counting with no cameras and no civil works, a radar solution may fit. uRAD Smart Traffic solution brings together models, capabilities, technical documentation and integration for real projects.
- No camerasCounting and classification with no personal imagery
- No roadworksMounts on a mast or pole, no road excavation
- 24/7Day and night, stable in rain or fog
uRAD Smart Traffic traffic radar
Frequently asked questions
Frequently asked questions about vehicle counting and classification
What is a vehicle counting and classification system?
It is a combination of sensors and software that detects road users passing a point on the road, counts them and groups them by type (pedestrian, bicycle, light, heavy), and can add speed, direction and lane. The data is used for mobility planning, traffic management and smart city projects.
What is the difference between vehicle counting and traffic surveying?
Counting focuses on counting and classifying vehicles. Surveying is a broader concept that also includes analysing volumes (AADT), hourly patterns, occupancy and speed to characterise how the road behaves over time.
What technologies are used for vehicle counting?
The most common are FMCW radar, cameras with video analytics, inductive loops, pneumatic tubes, manual counting and Bluetooth/WiFi sensors. Each offers a different trade-off between data, privacy, installation and maintenance.
What advantages does radar have over a camera?
The radar captures no personal images, works day and night and is more stable in rain or fog. Cameras add visual detail and licence plates, but involve image processing and depend on lighting. The choice depends on the project goal.
Can a camera-free system classify vehicles?
Yes. Technologies such as radar can classify by length or type (pedestrian, bicycle, light, heavy) without capturing personal images. The level of detail depends on the technology, the configuration and the conditions at the measurement point.
What data can be obtained in a counting project?
Depending on the technology: counting per lane, direction, speed, classification by type, annual average daily traffic, hourly patterns and occupancy, integrable into ITS, IoT or Smart City platforms.
When is radar preferable to inductive loops?
Radar tends to be preferable when you want to avoid civil works, measure speed and direction, work 24/7 without depending on lighting and reduce maintenance. Loops require opening and replacing the road surface, though they remain valid for permanent installations.
What factors affect counting accuracy?
The sensor position and height, road geometry, dense traffic, occlusions, static objects, mount vibration and weather. That is why calibrating and validating the data in the field is recommended.
Can the system integrate with Smart City platforms?
Yes. Many systems offer standard outputs and protocols to integrate with ITS, IoT and Smart City platforms, allowing data to be centralised and used for traffic management and urban planning.
What information do I need to request a technical assessment?
It helps to define the type of deployment, number of lanes and directions, whether classification and speed are needed, whether civil works are allowed, privacy requirements, power and connectivity, desired integration and environmental conditions. With that you can check your project’s feasibility.
Camera-free radar counting
Next step
Have a vehicle counting and classification project?
Whether you want to compare technologies or assess a specific road, you can review the radar solution for vehicle counting or share your case so we can study its technical feasibility.