All About Inductive Loops

Principles and Theory of Operation

    Loop detectors operate on the principle of inductance, the property of a wire or circuit element to "induce" currents in isolated but adjacent conductive media. A detector consists of an insulated electrical wire, placed on or below the road surface, attached to a signal amplifier, a power source, and other electronics. Driving an alternating current (normal operating frequency between 10kHz and 200kHz) through the wire generates an electromagnetic field around the loop. Any conductor, such as the engine of a car, which passes through the field will absorb electromagnetic energy and simultaneously decrease the inductance and  frequency of the loop. For most conventional installations, when the inductance or frequency changes a preset threshold in the actuate detector electronics, this indicates that a vehicle has been detected. Many factors determine loop inductance, including wire size, wire length, the number of turns, lead length, and insulation. 

Detector Configuration

As noted above, the elements of a detector include:

    The inductive loop is an insulated electrical wire, usually several meters to a side, with several turns. Loops are installed in a variety of shapes such as square, rectangle, diamond, circular and octagonal, though each configuration produces a different electromagnetic field. For instance, diamond loops reduce the probability of detecting vehicles in adjacent lanes. The pull-box, usually located adjacent to the road, houses the splices between the lead-in cable from the controller and the lead-in wires from the loop. Lead-in wires are usually shielded and twisted to eliminate disturbances from external electromagnetic fields, such as adjacent loops. The controller electronics, usually housed in a rugged cabinet in a safer more accessible location, detect, amplify, and process loop signals. The controller orchestrates loop operation and provides power. A typical controller can handle up to forty loops, though in practice will probably oversee far fewer.

    Inductive loops may be placed either on the road surface, or up to twenty inches or more into the pavement. While deep buried loops exhibit a longer life span, their electromagnetic field is weaker and detection becomes more difficult . Loop sensitivity, defined as the smallest change in inductance which will cause actuation, decreases around 5% for every inch into the pavement the loop is installed. Since the characteristic shape of a detection is not distorted, additional loop turns which emit a stronger loop field can compensate for pavement interference. Unless loops are installed during road construction, installation requires a saw cut, up to 10 mm wide, into the pavement. Unfortunately saw cuts have been found to undermine the structural stability of the pavement in some cases.

    Wire type also influences loop operation. Most inductive loops are formed by wrapping a single wire strand around the loop shape a prescribed number of times. Because these turns are spaced randomly, the electromagnetic field, and consequently detection results may vary from detector to detector. It has been established that a multi-conductor cable, which holds the loops in uniform proximity, will produce more accurate measurements. Installation of the multi-conductor necessitates a wider cut into the pavement than the single wire configuration. Ruggedized, weather resistant pre-formed loops are available which aid in uniformity, but also can be more difficult to install.

    Detectors operate in either the pulse or presence mode. Presence operation, often used with traffic signals, implies that detector output will remain "on" while a vehicle is over the loop. Pulsed detection requires the detector to generate a short pulse (e.g. 100 to 150 ms) every time a vehicle enters the loop, regardless of the actual departure of the detected vehicle.

    Many detectors today employ digital technologies which sense a change in the resonant frequency of a loop due to a decrease in inductance. Digital techniques allow more reliable and precise measurements than their analog counterparts. Some digital units incorporate advanced electronics such as self-tuning amplifiers, open-loop test functions, and automatic or remote reset capabilities. These features can significantly reduce detector maintenance costs and calls. The newest detectors can actually output the digitally sampled inductance "signature" of each vehicle, allowing the development of flexible signal processing software to add considerable more robustness than the hardwired set "threshold" detectors. 

Selection Criteria

Several parameters characterize the performance of loop detectors:

Response Time

    Defined as the time between the when inductance crosses the preset "threshold" due to the arrival or departure of a vehicle, and when this is indicated on the digital output side of the detector. A consistent and fast response time is crucial for accurate speed measurements. Response time is affected by vehicle size, speed, detector type, sensitivity, and wire type. Response time decreases with smaller vehicles, which have lower ground clearance and a shorter distance to the engine block and axle. Faster speeds tend to reduce response time, as does increased sensitivity.

Recovery Time

    The time required for a loop to return to normal operation after a period of sustained occupancy. Recovery time is particularly important for vehicle counting. Loop standards dictate that after a sustained occupancy of five minutes a detector return to at least 90% of the minimum sensitivity within one second after the zone of detection is vacated.

Installation and Maintenance Considerations

    For comprehensive surveillance of mainline routes, detector stations (possibly pairs) should be installed every 600 to 1200 meters. Loops should also be stationed around access and egress points, and at any locations where operational problems occur. Provisions should also be made for maintenance access. Different degrees of processing of raw loop data may be performed at the TMC or remotely in the field. Remote analysis requires additional processing hardware for each detector station, but less sophisticated communication links, since only the data relevant to highway surveillance is transmitted back to the TMC. In contrast, centralized analysis requires the transmission of large quantities of raw data and greater hardware requirements (storage and processing) at the TMC. If transmission lines are leased, which is often the case, data transmission costs can become cost prohibitive.

    Annual maintenance costs average around 10% of the original installation and capital cost, adjusted for inflation.

    Loop failure rates are strongly related to maintenance and installation procedures. Surveys of state DOTs indicate that failure rates vary significantly, ranging anywhere from three to fifteen percent per year. In practice, loop inspection procedures also vary substantially among DOTs. Loops may be inspected anywhere from one to twenty-five times per year. Highest inspection rates occur where loop operation is critical to the operation of other deployed traffic management systems. In many instances loop maintenance costs are sufficiently high that malfunctioning loops are replaced outright, without any diagnosis of the cause of failure. Because loops have been deployed extensively, consistent installation recommendations and primary causes of loop failure have been documented.

Installation

    Installation recommendations for effective and long lasting loop systems include:

Mechanical Failure

    Many factors contribute to physical loop failure. Pavement and sealant (of the saw cut) failure are commonly identified as the primary culprits. Pavement failure or deformation (cracking, rutting, potholes, or shoving) causes loop wires to be strained resulting in breakage, wire insulation wear, or the infiltration of foreign materials. Sharp bends in loop corners have also been found to cause problems, such that the insulation deteriorated or was broken.

    Sealant failure poses additional problems. Once the sealant fails, the loop may become exposed or foreign materials may infiltrate the cut. In many cases the loop was found to have floated to the top of the cut, either before the sealant could cure or because it remained plastic. Other common sources of loop failure include poor installation and maintenance procedures, damage from utility repair or construction, lightning surges, detuned amplifiers, and corroded splices or wires.

Data Malfunction

    Many sources of loop malfunction can produce erroneous detector data. These include stuck sensors, hanging (on or off), chattering, cross-talk, pulse breakup, and intermittent malfunction. Cross-talk involves the mutual coupling of magnetic fields that produces interaction between two or more detector units which are in the same cabinet or in close proximity to each other. Cross-talk results in erratic loop behavior and inaccurate detections. Pulse break-up involves gaps in detector actuation data, which may be incorrectly interpreted as different vehicles. As described below, many of the these problems can be corrected with data filters. 

Attainable Information

    Loop detectors supply several pieces of information about prevailing traffic conditions, including vehicle presence, flow, occupancy, and velocity. A good loop detector system is cited as accurate to within 5%. The accuracy and consistency of detector output is a strong function of installation and calibration procedures. For example, it is possible that detectors with different sensitivities longitudinally separated by thirty feet give occupancy data which differs by 40%. Loop detectors are also limited by their inability to detect stationary vehicles.

    Flow and occupancy may be extracted directly from loop data. Speed may be approximated from the data of a single detector using the fundamental theory of traffic flow:

flow = speed * density

where density is approximated from occupancy by:

density = occupancy * g

and

where K is a conversion factor. While it is possible to obtain reasonable speed estimates with this strategy, paired loops offer a more accurate approach. Velocity is calculated from the travel time between two loop detectors which are separated by a known distance. Accurately calibrated speed traps with loops of individual wire can expect to achieve measurement errors of 5-8 kph (3-5 mph) at low speeds and 16-19 kph (10-12 mph) at high speeds. Multi-conductor cable loops average errors about 0.3 kph (0.2 mph) at low speeds and 5-8 kph (3-5 mph) at high speeds.

    There are several considerations in speed trap design. For one, loop inductance is a strong function of vehicle speed. One study determined that a vehicle travelling at 20 mph produced a 3% inductance shift, while another at 80 mph yielded only a 1% inductance shift. Sensitivity settings may have to be adjusted when ILDs are used in high-speed freeway environments. The separation between loops is another relevant variable. In practice anywhere from 2 meters to more than 20 meters is feasible. However if detectors are too close cross-talk may occur, while detectors spaced too far apart may be susceptible to vehicle lane changes. Suggested optimal spacing is around 9m. 

Data Reliability

    For traffic management strategies such as incident detection to be effective loop data must be reliable and accurate. Many TMCs use a combination of manual inspection and reliability tests to validate incoming data. Such tests serve a dual purpose, they flag erroneous data and identify malfunctioning loops. Various approaches are employed to identify inaccurate data.

    Initial error detection often occurs in the field. The data may be filtered, where pulses or gaps in actuation less than some brief interval, say one-fifteenth of a second, are ignored. The data may be flagged as unreliable if a microprocessor sees more than two valid pulses (vehicle endings) in a second. These tests usually detect gross errors, but other malfunctions may go unnoticed.

    More advanced filtering techniques are available to validate loop data. One approach is to compare a detector's on time to the average on-time of all other detectors at that station. A second strategy compares detector data (volume, occupancy, and speed) against realistic thresholds at periodic time intervals. For example, detector data is flagged if occupancy exceeds a predefined maximum for a certain period of time (say more than 90% for five minutes). A more complex algorithm uses a multi-regime comparison of the flow - occupancy ratio to maximum and minimum expected speeds. Occupancy is converted to density using a variable g which varies as function of occupancy. Research has shown that a constant g can introduce significant error into speed estimates. These algorithms achieve good detection rates with low false alarms rates, and often identify malfunctioning detectors overlooked by manual inspection. In practice most TMCs operate in a hybrid fashion, using several elements from the tests described above.

    New detector cards which are becoming commercially available can directly output the (digitally sampled) change in inductance. This allows the development of analysis software with considerable more robustness. For example, conventional "threshold" detectors may double (or even triple) count a long truck with many distinct changes in inductive mass. Conventional detectors may double count a vehicle changing lanes between loops, or not count them at all, depending on how sensitive each lane's "threshold" is set. If the base inductance of a loop changes due to mechanical wear or weather induced corrosion, these old threshold detectors must be manually re-calibrated in the filed. With the new cards, it appears plausible to develop software to mitigate these types of problems by intelligently interpreting any and all changes in inductance across all the roadway loops at one time. Unfortunately, this capability has not been operationally proven in this country at this time (as of early 1997). 

This is a secured server site. Your data is safely encrypted and is safe from unauthorized access.