0
\$\begingroup\$

I am having trouble locating solutions for a long range trigger sensor. Want to detect an approaching vehicle 300 feet or more down the road from a pole mounted device. My first thought was a laser or IR beam break sensor, but the only ones I have seen need a reflector at the target point and the unit registers the absence of the reflected beam when obstructed. I cannot put a reflector in the middle of the road.

Only device I have ever seen that works similar to what I need are laser range finders but I would have no idea how to reverse engineer one. This sensor would supplement a PIR dome lens sensor in a security application. Can't give too many specifics but the PIR works great for pedestrians. However there is an inherent delay within my system between trigger time and when the system becomes active, meaning that fast moving vehicles are long gone before the system activates.

\$\endgroup\$
  • \$\begingroup\$ How often do vehicles arrive? How fast are they going? Any possibility of a vehicle approaching while another is leaving on the same road? Do you care about motorcycles or bicycles? What modifications can you make to the roadway itself? How much access do you have to the area adjacent to the road? \$\endgroup\$ – Elliot Alderson Aug 2 '18 at 19:30
  • 1
    \$\begingroup\$ This is either a shopping question or it's unclear what you're asking. Suggest you clarify what the specific Electrical Engineering problem you're trying to solve is. \$\endgroup\$ – brhans Aug 2 '18 at 19:30
  • \$\begingroup\$ Do not have any pictures handy, and yes this is sort of a shopping question. I was hoping the community of Electrical Engineers might be more familiar with a product that could work or a sensor I could adapt. What I am basically working on is an LED security light that is activated by pedestrians and passing cars. It doubles as a street light so it needs to be activated a hundred or more feet ahead of incoming vehicles. Biggest problem is the light fixture I am using has a "soft-start" function that I cannot deactivate so it takes about 2 seconds to come up to full brightness. \$\endgroup\$ – Jason P Aug 2 '18 at 19:56
  • 1
    \$\begingroup\$ Use images processing and a video camera. It's 2018 outside. \$\endgroup\$ – Gregory Kornblum Aug 3 '18 at 4:43
  • 1
    \$\begingroup\$ Vision systems may not work in the dark at 100m. But if you can’t afford a 100m phone cable installation you might not afford a more sophisticated solution \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Aug 3 '18 at 4:45
1
\$\begingroup\$

One idea: Use a CW microwave radar motion detector (=an intrusion alarm device). Place the sensor in front of a big enough parabolic reflector to get narrow enough transmitting beam and big enough echo signal catching area. You must find the needed reflector size and focal length by making experiments.

There's a couple of traps: Vehicle speed can be out of the detection range (=too high doppler shift) and your local law can limit the allowed beam power densities. Detector's transmitter can be legal as is, but when all power is directed to a narrow beam, the beam can be illegally strong.

\$\endgroup\$
0
\$\begingroup\$

Another possibility, if you (or your customer) owns the road in question is to install an electromagnetic vehicle detector. You install a single conductor cable mechanically protected under the surface of the roadway in a loop about the size of a compact car's outline or a bit smaller. Apply an AC signal and measure the current. A car being present above the loop will act as an iron/lower reluctance core and increase the inductance of the loop, affecting current.

If you see a square pattern on a roadway with cuts across the corners before an intersection, this is where a sensor has been cut in after the fact. A double bladed concrete saw cuts a trench for the conductor, with the corners cut to avoid bending the somewhat rigid cable too harshly, the cable is laid into the trench and the gaps are patched with asphalt/tar/cement as appropriate.

You may also be able to simply mount a second PIR sensor at the side of the roadway and at an upstream location. If you have networked lighting controls, this could be quite viable, cheap and easy. Even without it may be a good option. There is a wide range of available sensitivity, and you can get their trigger area down to a fairly small cone or wedge with some types.

\$\endgroup\$
0
\$\begingroup\$

Take a LED (array if needed) producing 1 watt of IR. Run that thru a lense system to generate 6" diameter beam of low dispersion. Aim that beam down the road, or across the road as some slight angle. Modulate the beam at 30kHz, so TV remote controls are not a problem. This is your transmitter.

Use a 1" lense, focused on a photodiode, to monitor that portion of road 300' away. This is your receiver. Assume the Receiver is 200 feet from the intersection of the TX beam and the RX beam.

When the car intersects the TX beam, that 1 watt of IR energy at 30kHz will be scattered; assume the scattering is uniform over a hemisphere; we now have an energy density to use for some math.

The RX lense, 1" diameter, is 200 feet from the car. The area of a hemisphere 200 feet in radius is 0.5 * (area of sphere with radius 200') or 0.5 * 4* PI * 200*200, or 6.28 * 40,000 or 250,000 square feet. In square inches, to easily work with our RX lense, we scale 250,000 by 144 (inch/foot)^2 to 40,000,000 square inches.

Thus we've just computed the RX system will have energy of 1 watt / 40Million or 25 nanoWatts of energy. Modulated at 30kHz.

Now you have a simple sensor problem. Assume the car is in the beam for 10 milliSeconds, so you know a noise bandwidth. Power is 25 nanoWatts.

\$\endgroup\$
0
\$\begingroup\$

I think you've already said a beam break solution is not going to work, so it was perhaps a poor title for your question.

If you want to detect an object at 300ft then the arc posed by the target object is quite low (around 1deg for a car front view) and you need a sensor with equally high angular resolution.

I'd suggest you might experiment with something like this Garman Lidar unit (good out to 120ft) which can give you a 3D point in space detector. This would be as close to 'beam-break' as you could get with a single non-moving sensor. There are quite a number of single plane/point Lidar solutions you can investigate.

If you want to experiment with lower resolution (area saturation) sensors you could try the TerraRanger EVO range, with the new EVO 60M a possibility. You can get these in the US from Robotshop. I doubt this is going to work at your required range though.

If this is a professional solution then there are lots of Lidar units available that will provide the required resolution and range ...but they are professional prices too. This Ibeo automotive Lidar would fit the bill, or perhaps this SICK LD-MRS unit would be ideal, both of these are 4 plane Lidar. Autonomous Stuff carries a wide range of Lidar.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.