Author: Hector López García
Why should we continue using invasive techniques to detect the number of axles and size of the wheels of a vehicle? The short answer is simply, We shouldnt!
The need to count vehicle axles and measure the size of the wheels has become an important feature for a number of traffic statistic applications and automatic vehicle classification (AVC) solutions.
The most common systems for detecting these characteristics are based on sensors installed in the roadway. The asphalt is cut to embed pressure sensors that every vehicle wheel touches. These sensors suffer much degradation and an increasingly lower performance in a relatively short time; more so on roads with heavy traffic. Nevertheless, it is easy to observe on motorways or city ring roads just how many of these sensors are in the roads, indicating their continued popularity.
A further step in wheel and axle detection technology is the integration of optical sensors in the AVC systems. Similar to a laparoscopic intervention on the road, the optical sensors are introduced on the lateral side of the road lanes, avoiding a material intrusion on the road surface. Instead of having sensors crossing the road, in this new situation we find light beams, invisible to the human eye.
The benefits of this approach are soon apparent: there is no sensor degradation caused by the passage of the vehicles, so the road can get up-to-date with its periodic maintenance works of rebuilding, fixing or salt spreading without affecting the sensors.
The performance of the pressure sensors also improves as the optical sensors greatly help to eliminate false wheel or axle detection.
With these types of AVCs, a new functionality is provided: the ability to detect the raised axles found on heavy vehicles, such as trucks and big buses. With road pressure sensors, wheels and axles can only be detected when they are touching the ground. It is not possible to deploy systems that detect raised axles only using pressure sensors; other devices need to be added.
Although the weakest point of optical detection technology is traditionally considered its robustness against extreme environmental conditions such as heavy rain and mud, new developments in LEDs and lenses combined with photocell redundancy integration make an overall system more tolerant to all environmental conditions and improve the performance of AVCs. This is particularly true for roads with ice on them; pressure detectors can suffer from the freezing and from contact with salt, whereas the optical detectors perform better.
All of these points have encouraged several road operators to deploy AVC systems equipped with optical sensors, thereby increasing their degree of satisfaction in performance and long-term maintenance. These solutions have been successfully deployed in countries with vastly different weather conditions, such as Mexico, Brazil, Chile, India, Russia and Spain.
An AVC system performs best when it uses optical sensors with customized logic and algorithms for a particular countrys requirements. The technology to do this is available for most of the current AVC traffic applications.
Whats next?
At the moment, optical detectors are a good solution when detecting and counting wheels and axles. However, more wheel and axle features can be extracted if pattern recognition conducted on video cameras is also incorporated. This technology is still immature in the field, as the algorithms that need to recognize wheels and axles in all environmental conditions are not fully developed for each type of application. In any case, there is good progress happening and much more to come from this sector in the future.
Need to know?
Vehicle detection tools have already improved greatly; and theres much more to come
> Automatic detection and classification systems allow the identification of vehicles through the measurement of physical parameters, with no need for human intervention, for the purposes of auditing, revenue collection management, statistics generation, tolling, etc
Tecsidel provides a range of detection and classification systems, including AVC systems with optical sensors