Hello,
I am looking for a range finding sensor in smoky environment, with the accuracy is ~1mm at a distance ~ 1.5m.
I foud a laser one that worked quite well but when smoke comes, we got higher error (could get even 20mm wrong)
A stereo camera may work better with smoke (we still can see objects in smoke quite clear so i think matching keypoints can be found). However, stereo can’t achieve 1mm accuracy.
I heard sonar can work well with smoke also but couldn’t find any accurate one.
Any advices are appreciated
Dear son_quang.
Thanks for your question! As you mentioned, each type of sensor has its strengths and limitations, particularly in smoky conditions. I will connect you with some of our suppliers who have expertise in this area and may be able to direct you to approprieate sensor. Hope it helps!
Hello son_quang, Here is a paper with a conceptual design for a simple optical rangefinder. This one is 3 mm resolution increment at 1.5 meters but it would be straightforward to change the laser and get 1 mm resolution. The design is flexible enough that you’ll be able to select a wavelength that has the best penetration through the smoke. If I had to guess, based on analogies with tissue optics, I’d say this is likely to be in the near-IR (NIR), or short-wave IR (SWIR), between say 800 - 1400 nm as the Rayleigh and Mie scattering vary as a high inverse power of wavelength. If you can work at 1300 nm you’ll be able to take advantage of an enormous investment by the telecom industry in optical components. After that the performance will be determined by how reflective your sample is at the chosen wavelength. Can you add a fiducial marker to the target to increase contrast? Hope this helps, John
“Range finding using frequency-modulated laser diode” Glenn Beheim and Klaus Fritsch 1 May 1986 / Vol. 25, No. 9 / APPLIED OPTICS 1439
Hi John,
Thank you very much.
in fact the laser we tested is with red signal (i dont know exactly the wavelength but it is visible, so i guess it a bit less than IR).
Do you know if there are any companies/services can design it ?
Hi Son,
A red laser will be more strongly scattered than a NIR / SWIR laser, so the signal attenuation and multiple scattering trajectories from the smoke particles are probably contributing to the range error in the smoky environment. NIR / SWIR lasers are less strongly scattered and the scattering is more in the forward direction, so the pathlength blurring will be reduced.
I’m not sure if there are commercial sources for something like this. It might be worth looking at the automotive LiDAR companies for off-the-shelf modules? Yes - there are companies who can design and build you a one-off unit, but it could get expensive? Do you have a budget in mind. Do you need a well-finished unit with all the IEC / UL / Regulatory clearances, or can you use a large, ugly, but functional benchtop prototype at first! I can build you the latter, and introduce you to people who can do a well-finished commercial product. John
Maybe it’s a wild idea, but you could use interferometry. Use a 1500nm (or so) laser diode beam which would have low scattering. Maybe a simple Michelson interferometer would do: you split the beam into two parts (maybe 95:5 ratio). One part (reference arm, 5%) you circulate in a reference arm or fiber 1.5m long. The other arm of the interferometer would be the measurement distance. You move the reference arm and record the changing minima/maxima on a InGaAs photodetector until you find the best contrast. Similarly to an OCT, just work on a larger distance.
Hi John,
We conducted a test with another red-light laser, but much stronger (i.e 200m range) and it worked with much smaller noise in smoky environment (still oscillate ~2mm but much mostly stable) .
Now we are looking for its digital version to read continuously and auto transfer data to our embedded computer.
Hi Son, That sounds plausible. The signal-to-noise ratio of an optical signal detected by a photodiode is given by N/2B where N is the photoelectron generation rate and B is the system bandwidth. So for the same bandwidth electronics, an increase in laser power gives an increase in the photoelectron generation rate N, so an increase in SNR, which could translate into a more accurate measurement. The residual range error could be blur associated with scattering of the laser light from the smoke particles, which has a very distinct intensity and angle dependence for a given wavelength and particle size.