5

I am currently thinking about making a robot that will autonomously drive around the place. The place I want this robot to drive in however contains quite a few glass walls. When mapping the area I would need to be able to see the glass. For this reason I am in need of a sensor that can see the glass, and not see through it. What kind of sensor would be the best for me? I need it to have a maximum range of about 2-10 meters and a minimum range of about 0.25 meters (preferably as small as possible). I was thinking about maybe using ultrasonic, but I was told that a laser-based sensor would probably be best. I could however only find industrial grade laser sensors that could see glass/transparent objects.

John Bow
  • 53
  • 1
  • 5
  • Thermal sensors, while expensive see glass as an opaque surface. You might get similar results using infrared lamps and detectors. – Octopus Mar 08 '17 at 06:34
  • @Octopus considering it will be on a moving robot in a regular environment, I doubt that using infrared lamps and detectors would work. Having to put up detectors all over the place is kind of out of the question. I will however have a look at thermal sensors. Thanks for the input! – John Bow Mar 08 '17 at 12:15
  • @Octopus I can't seem to find any information on Thermal Sensors that can tell me how far away the glass would be. I might have missed mentioning this in the post, but to avoid collisions with the glass, I ofcourse need to know the distance between the robot and the glass. – John Bow Mar 08 '17 at 12:31
  • Sorry, I didn't mean to imply that a thermal sensor can be used to measure range, it would be too complex and expensive anyhow, but that since thermal sensors see glass as opaque, then I imagine a basic IR proximity sensor probably would, too. – Octopus Mar 08 '17 at 17:56

3 Answers3

3

Unless you need very good resolution or a very tight detection pattern, I would go with an ultrasonic sensor. They can be very cheap and easily meet your detection range specs, for the very low end look at something like the HC-SR04, and for more serious applications Maxbotix offers a wide range of sensors.

Mark Omo
  • 1,919
  • 13
  • 23
  • Thankyou. This is very useful. Do you maybe have some examples of laser-based sensors too? So I can compare them? – John Bow Mar 08 '17 at 12:19
  • 1
    Sure, the biggest one that comes to mind is the Lidar Lite V3 but it cannot detect surfaces that are reflective or clear, same for all LIDAR units that I know of in the lower than $1200 range. For the 0-10m range I think that Ultrasonic sensors are best, and for less than that IR proximity sensors will work well. – Mark Omo Mar 08 '17 at 20:48
3

I'm pretty sure that a very basic IR proximity sensor would do the trick. Glass is opaque to all but visible light. IR (as well as UV) will not penetrate the glass and you ought to be getting reliable distance measurements to the glass's surface.

Octopus
  • 1,047
  • 1
  • 6
  • 13
  • Aha, this is what I was looking for. I'll have to look around for one with a decent range, since I need it to be able to measure to at least 2 meters from the sensor, but that will probably be doable just fine. Shame a LIDAR won't work. Since that uses laser instead of IR right? LIDAR seems very appropriate for mapping, just not when glass is involved. – John Bow Mar 09 '17 at 09:46
  • This was indeed the answer I was looking for, since I needed another option to compare my default option of an ultrasonic sensor with. I will however in the end use an ultrasonic sensor. This is due to the issues Infrared sensors can have with sunlight, and the fact that ultrasonic sensors generally have a longer range. Thanks again for your input! – John Bow Mar 09 '17 at 11:14
  • " Glass is opaque to all but visible light." this is very incorrect. https://www.google.com/search?q=glass+transmissivity – Damien Sep 27 '21 at 04:15
  • @Damien, within the scope of the question, my comment is valid. Is there something in a linked article that would make my comment misleading within this context? It wasn't an engineer's question. I mean, if you want to get technical, we could argue about what "opaque" means. – Octopus Oct 12 '21 at 21:57
  • IR will get through the glass as well as visible light (~400-750nm), up to at least 2500nm, which is way beyond any IR optical sensor, which makes your answer incorrect and misleading. There are other ways of detecting glass through light polarization, intensity, diffraction, and so forth. A generic sensor "might do the trick" but will yield an unreliable system inherent to the environment (lighting, background, etc..) if that sensor wasn't specifically designed for that purpose. – Damien Oct 13 '21 at 08:26
  • @Damien, I know from experience that FLIR cameras do not see through regular glass. Special glasses exist that do allow IR to pass through, for example the lens that took this picture obviously does. picture – Octopus Oct 13 '21 at 17:41
  • @Octopus FLIR operates in the 8000–15000 nm range, this is very deep IR and optical sensors for object detection don't operate in this range, unless very specific and very expensive. Typical IR optical sensors are in the 700-980nm range, to which glass is transparent. – Damien Oct 14 '21 at 08:00
0

Do you maybe have some examples of laser-based sensors too?

I will recommend this Mini Lidar sensor that only costs $39.9

Also,you do need the ultrasonic distance sensor

Measurement range schematic diagram

Elaine0228
  • 1