Author: Aidan Smyth, electronics engineering technician, has worked with the NRA for the past 20 years and has primary responsibility for the installation, maintenance and repair of the NRA’s Automatic Traffic Counter systems The concept of monitoring road traffic is nothing new. In Ireland, significant traffic surveys were carried out by local authorities as far back as 1925 when the bulk of traffic would have been horse-drawn vehicles. Times have moved on and vehicle traffic on our roads has increased hugely. In line with that, so too has the demand for traffic volume data. Traffic data is required and used by many parties including the NRA itself, local authorities, consultants, road engineers, students and, of course, the general public.

NRA traffic monitoring site with renewable power supply[/caption] Data is used for many purposes, not least road building and planning, road safety and environmental aspects and more commonly now as an economic indicator. The National Roads Authority (NRA), which has overall responsibility for the national primary and secondary road network in Ireland, has been monitoring and counting traffic since its inception in 1994 and prior to that in its previous guises as the Environmental Research Unit (ERU) (1980s) and An Foras Forbatha Teoranta (1960s/'70s). The NRA currently operates and maintains more than 270 Automatic Traffic Counter sites across the national primary and secondary road network.

Inductive loop detector

For its permanent traffic counting stations, the NRA uses inductive loop detectors. The inductive loop detector has been used as a traffic detection sensor since the 1950s. It is relatively simple to install and has a long shelf life. Advances in electronics and associated software engineering have led to more reliable loop detection systems, allowing for accurate count and relatively accurate vehicle classification using just two loops per lane. We all correctly associate roads with civil engineering, so people don’t generally think of roads containing electronic sensors. Many will be familiar with loops and have an idea how they work but I am still frequently asked, 'What exactly is an inductive loop detector?' 'How do loops work?' 'How do loops classify vehicles?', or some variation on the same. Almost always because there is a sensor buried in the road, there is a pre-emptive assumption that there is 'pressure' on the sensor so vehicles are being weighed. Not so. Loop detectors do not measure weight nor does a standard loop count actual axles. So what is a loop and how does a loop detection system work? For the purposes of counting traffic, an inductive loop detector, is little more than a coil of wire buried in shallow (75-100mm) slots cut in the road surface. Slots are normally cut in a square shape, approximately 2m x 2m with three turns of wire 'looped' into the cuts. The tails or feeder wires are twisted and fed back to electronics at the road edge.

Inductance, resistance and insulation resistance

The loop has fundamental electrical properties which are measured to ensure suitability for purpose, namely, inductance, resistance and insulation resistance. Loops use the principle of electromagnetic induction, i.e., when a current flows in a coil of wire, a magnetic field is 'induced' around the wire. (See Figure 1.) Figure 1: Magnetic field induced by current flowing in wire The roadside electronics (counter) applies a medium frequency electrical signal (between 30KHz and 150KHz) to the loop, thereby creating a magnetic field around the wire loop. The signal frequency is steady, so the road loop now becomes an integral component (inductor) of a 'tuned' electronic circuit. As long as the signal is applied, the magnetic field remains around the loop. The magnetic field is approximately half the height of the longest side so, for a 2m loop, the field will be c.1m in height. Now, when this magnetic field is disturbed by metal, i.e., a vehicle passing over the loop, the metal absorbs part of the magnetic field energy, inducing changes in the magnetic field and the frequency of the applied signal. The inductance of the loop also changes. These changes are all induced, hence the term, 'inductive loop'. As a result of the vehicle crossing over the loop, the circuit 'detunes' momentarily and, when the vehicle has passed, the circuit returns to a steady tuned state. The electronics detect and measure the changes in frequency and/or magnetic field and/or inductance, so thanks to the principle of electromagnetic induction we now have a 'metal detector'.

Basic traffic counting

The electronics records each metal detection as a vehicle count, so by placing a loop in each lane we can get a basic directional volume count but little else. The system can measure the length of time the loop is detuned as a vehicle passes over, i.e., the loop 'on-time', but this cannot determine vehicle class or vehicle speed. A long 'on-time' could represent a short vehicle moving slowly or a long vehicle moving quickly. A Multiple Lane Traffic Counting Site. The car is directly over the first “Loop” and approaching the second.[/caption] Therefore a second loop is added, about 2.5 to 3m from the first. The leading edge distance c. 4.5m is programmed into the electronics and by measuring the time lag between first and second loops detuning, the system can now also measure speed (distance/time).

The system measures both individual loop on-times and mutual on-time, i.e., time both loops are activated simultaneously as the vehicle passes over. The counter manufacturer uses these measurements and applies mathematical algorithms, within the system software to calculate the length of a given vehicle. Now we have a vehicle classification system based on length. But again it doesn’t stop there. Figure 2: Loop magnetic field If we take figure 2 as a typical loop magnetic field then we can see that a lower chassis vehicle crosses more lines of magnetic flux than a higher chassis vehicle, therefore the lower chassis vehicle affects the magnetic field (and loop frequency)  more than higher chassis vehicle. So, contrary to what one might expect, a car with a lower chassis affects the loop more than a larger higher chassis truck. Different vehicle types or classes will therefore have different detune effects. And, likewise, similar vehicle types will have similar detune effects. These loop detune changes are very real changes and can actually be graphed against time. Figure 3 shows the difference in typical graphed detune results for different vehicles. One can see the individual on-times as well as the mutual on-time for the loops.

The manufacturer can now develop sets of parameter figures relevant to various vehicle types using the profiles caused by the changes. The counter is then pre-programmed with these sets of parameter figures for each profile (vehicle type) and as various vehicles pass, the system measures the changes, compares and associates the changes and measurements with a given parameter set and then bins the vehicle into the relevant group depending on where the cut-off points for various vehicle types are.

Loop classification system

Ultimately we now have a system capable of classifying vehicles based on 'profiles' determined by induced changes in the loop properties combined with length measurements. The more accurately the system can measure these profile changes and vehicle lengths, the more accurately the system can determine different vehicle classes. This is what is commonly referred to as a loop profiler or loop classification system. Accuracy levels for this type of system varies from manufacturer to manufacturer and across vehicles being classified but can reach >98 per cent for cars,  > 90 per cent for LGVs (light goods vehicles) and > 85 per cent for HGVs (heavy goods vehicles). We do not need to know exactly every type of vehicle out there so, in general, the NRA classify about seven to 12 different vehicle classes, which is sufficient to get an overall picture of vehicles types. In the NRA’s case we have a dedicated interactive website which presents the traffic data retrieved from individual Automatic Traffic Counter sites in various formats. The data is freely available at https://www.nratrafficdata.ie. Aidan Smyth is an electronics engineering technician and has worked with the NRA for the past 20 years and has primary responsibility for the installation, maintenance and repair of the NRA’s Automatic Traffic Counter systems