Sensor Evaluation

Traditional radio frequency (RF) radar systems use a different technique to directly measure the speed of a target. The Doppler effect is employed to ascertain the speed of a moving object. RF energy is slightly altered when reflected from a moving target. The frequency of the reflected signal is lower when the target is moving away from the receiving station and higher when moving toward the station. Careful analysis of the frequency difference can reveal the speed and direction of travel of the target with respect to the receiver. Being that the microwave emissions from the radar are much lower in frequency than that of light, more of the incident signal is reflected back to the receiver by the sides of metallic railcars. Testing demonstrated that it is relatively easy to acquire speed information of the head end of a train and the following cars.

The sensor technology adopted for the off right-of-way detectors was RF Doppler radar. The choice enabled several important concepts. RF radar systems employ a mature technology. Research and development on microwave components has been underway since the 1930’s and solid state equipment is now common and reliable. Motion sensing radar products are widely used as triggers for door or gate openers, intrusion systems and as precision speed sensors. Because of advances in solid state integrated circuitry, the transmitter / receiver can be physically small and lightweight. A Doppler system will work adequately in pole mounted side-fire installations or overhead, aimed directly at the target as long as the appropriate off-angle corrections are understood. Range issues should easily be met with Federal Communications Commission (FCC) part 15 equipment, insuring low power consumption and lower cost. RF radar systems can offer presence (or range) as well as speed and direction detection if desired.

Radar unit under evaluation

A search was conducted to identify potential suppliers of an applicable radar product. A review of current roadway and police radar detectors revealed three companies with products that were worthy of further study. Each company was contacted. Discussions with the first vendor revealed that one company was preparing to introduce a radar sensor that would offer real-time speed data. A pair of prototype traffic Doppler radar sensors was offered for evaluation. The sensors provided an ASCII representation of the speed of a target via a RS232 serial data port. Unfortunately the direction of the target could not be recovered from the unit as designed. An internal switch determines the direction of movement that the sensor observes and is designed to be set during installation. Subsequent discussions with Microwave Sensors resulted in an agreement for TTI to modify the unit to enable remote direction change capability. Tests conducted on the sensors were successful. The sensors were small, lightweight and consumed little power during operation. The RF power output was easily sufficient to detect trains at our required range. The speed measurement accuracy was tested using a laser radar speed reference. The RF unit showed close agreement in speed detection accuracy.

The second vendor was contacted and a detector was purchased. Their sensor was designed to be mounted over a lane of traffic or on a pole mounted off to the side of a traffic lane. The sensor offered an RS232 serial data port and an ASCII representation of vehicle relative speed. The unit was larger in size than the first vendor's product and likewise can only detect movement in only one direction, configurable via an internal jumper. The procured unit was never fully tested due to malfunctioning electronics on the system processor board.

The final candidate product came from a police radar vendor. The company recently provided speed radar units to Burlington Northern Santa Fe Railroad's Argentine Yard upgrade project in Kansas City. A modified police radar was delivered and tested. The unit, like the others, was small, low power and provided a serial data link. This unit was the only one of the group that was able to determine movement direction. Field tests revealed that though the sensor easily measured passing trains, background objects or noise produced spurious train detects.


After evaluating available sensors and gaining experience with radar equipment in general, a list of requirements was derived for the optimum rail sensor for the project.

• Speed range from 1 MPH to 70 MPH (unless the system is to operate in high speed rail corridors)
• Speed resolution of 1 mile per hour with accuracy of 1 MPH
• Speed updates at least once per second
• Provide the direction of the moving target
• ASCII serial data stream at 9600 bits per second
• Ability of the sensor to automatically deliver a speed reading or be polled (allowing a control unit to access data when desired)
• Relected signal strength measurement useful for target identification
• Provide presence detection (desirable but not a requirement)

Sensors Arrive and Undergo Testing

Click here for information of the test setup.


Graph showing a typical outbound train passing the sensor.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Graph showing an elevated noise floor - possibly due to rain at the site but not verified.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Typical outbound train.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Inbound train.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Data taken from the radar at approximately 5:12PM. Notice the preponderance of outbound "clutter". The clutter is evening peak traffic on a distant roadway.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Outbound unit coal train. The train is made up of over 100 identical open top aluminum hopper cars.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Inbound general merchandize train. Notice that the radar begins detecting the train's speed many seconds before it arrives at the site (large jump in signal strength).

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Data sample from midnight illustrating a low noise floor.

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div

Relative Power 10dB/Div Speed 10MPH/Div Radar Reflected Signal Profile Legend Inbound Speed Outbound Speed Signal Strength Time --- 25 seconds / Div