high speed, narrow beam RF reflection detector

Thread Starter

Corey Haddad

Joined Oct 14, 2015
22
Imagine I had a laser pointing straight down from a car, and a detector to see when the beam was reflected back up. Now suppose the detector samples at something like 1 MHz. If I were driving down the highway at speed, and I put a small mirror on the road surface under the path of my car and laser, I would have many 'True' readings as I pass over the mirror.

What would be the simplest approach to having something similar using RF of some type? Instead of an optical mirror, I would use a thin strip of metal as a reflector. It would be just 1 cm wide in the direction of car travel (but could be a meter in length perpendicular to said direction). It could be on the surface of the pavement, or ideally even 1 or 2 cm beneath.

If this works the way I would like, if one knew the speed of the car, one could measure the width of the metallic strip by looking at the time stamps from when the RF signal was reflected and when it wasn't. The overall distance between the bottom of my sensor and the reflective surface would something like 0.5m - 1m.

• I could imagine using some kind of mask to block the RF except for that in the shape of narrow beam
• Microwaves of mm waves might be better than longer waves?
• SnR consideration: I am just attempting to detect the presence of a reflector, and not trying to pass data
• The beam needs to be narrow in the direction of travel, but it could be wide in the perpendicular direction

Any ideas for how feasible this would be to make or what kind of technologies would be applicable?

Thanks!
 

drjohsmith

Joined Dec 13, 2021
852
Imagine I had a laser pointing straight down from a car, and a detector to see when the beam was reflected back up. Now suppose the detector samples at something like 1 MHz. If I were driving down the highway at speed, and I put a small mirror on the road surface under the path of my car and laser, I would have many 'True' readings as I pass over the mirror.

What would be the simplest approach to having something similar using RF of some type? Instead of an optical mirror, I would use a thin strip of metal as a reflector. It would be just 1 cm wide in the direction of car travel (but could be a meter in length perpendicular to said direction). It could be on the surface of the pavement, or ideally even 1 or 2 cm beneath.

If this works the way I would like, if one knew the speed of the car, one could measure the width of the metallic strip by looking at the time stamps from when the RF signal was reflected and when it wasn't. The overall distance between the bottom of my sensor and the reflective surface would something like 0.5m - 1m.

• I could imagine using some kind of mask to block the RF except for that in the shape of narrow beam
• Microwaves of mm waves might be better than longer waves?
• SnR consideration: I am just attempting to detect the presence of a reflector, and not trying to pass data
• The beam needs to be narrow in the direction of travel, but it could be wide in the perpendicular direction

Any ideas for how feasible this would be to make or what kind of technologies would be applicable?

Thanks!
why does beam need tp be narrow,
make your reflector resonant at the frequency you want
then the reflection will be great, whilst reflection off the so-rounding would be small.
 

Thread Starter

Corey Haddad

Joined Oct 14, 2015
22
Nice! I hadn't considered the resonance of the reflector.
I was thinking the beam would need to be narrow so it would be quite clear when my sensor was passing over the reflector with a high degree of spatial resolution. Imagine I had two reflectors, each 1cm wide with 3cm spacing between them. I would like to be able to detect each individually, and the spacing between them (again, giving I know the speed of my sensor and sample rate).

Upon doing some reading (ChatGPT), it seems like ground penetrating radar techniques could be applicable, but I am curious about a simpler system since I am only trying to see if my chosen reflector is present or not.
 

Ya’akov

Joined Jan 27, 2019
9,070
Upon doing some reading (ChatGPT)
It is a mistake to use ChatGPT as a reference, it’s not one and not intended to be one. It “lies” and “hallucinates” and it’s main ability is being able to produce plausible explanations without particular regard for the truth.

If you really want to use ChatGPT, use Bing which uses ChatGPT but also lets it search the web and focus it on factuality rather than plausibility.
 

Janis59

Joined Aug 21, 2017
1,834
About "best friend of humankind"- the Chatbot GPT4. The workmate sitting just behind me asket chatbot what is wrong with he having specified defect of 594 PWM tablet. GPT answered - the problem is hidden in feet number 17 connected components. He replied back: "but... the tablet has only 16 feet on the body". GPT bit waited, then answered - (how polite and nicely correct!) - yes, I have mistaken. For the truth, the problem is hidden in components attached to feet number 18. :) :) :)
 

drjohsmith

Joined Dec 13, 2021
852
Nice! I hadn't considered the resonance of the reflector.
I was thinking the beam would need to be narrow so it would be quite clear when my sensor was passing over the reflector with a high degree of spatial resolution. Imagine I had two reflectors, each 1cm wide with 3cm spacing between them. I would like to be able to detect each individually, and the spacing between them (again, giving I know the speed of my sensor and sample rate).

Upon doing some reading (ChatGPT), it seems like ground penetrating radar techniques could be applicable, but I am curious about a simpler system since I am only trying to see if my chosen reflector is present or not.
GPR,assumes the subject is effectively static /not moving,
your detecting a change,QED moving.

Will your object be attenuating the "beam"
if so , with two resonant detectors, the return will drop when first sensor is covered, then drop more when second sensor covered.
 

Thread Starter

Corey Haddad

Joined Oct 14, 2015
22
drjohnsmith -
The reflectors would be fixed in place, and could be either on the road surface or very slightly below it. The would be metallic strips running perpendicular to the direction of car - and therefore sensor - travel. Ideally my sensor is looking straight down, with a narrow FOV, with good resolving power. It would be nice to be able to detect passing over a strip 1cm in width, and to be able to detect two such strips if separated by a distance of 4cm or so. Sub-cm resolving power would be great, but seems unrealistic and isn't strictly necessary. And again - consider the separation between the sensor and the strip(s) to be 0.5 - 1 m.

(Discussion of ChatGPT removed by moderator because it is off topic here.)
 
Last edited by a moderator:

drjohsmith

Joined Dec 13, 2021
852
drjohnsmith -
The reflectors would be fixed in place, and could be either on the road surface or very slightly below it. The would be metallic strips running perpendicular to the direction of car - and therefore sensor - travel. Ideally my sensor is looking straight down, with a narrow FOV, with good resolving power. It would be nice to be able to detect passing over a strip 1cm in width, and to be able to detect two such strips if separated by a distance of 4cm or so. Sub-cm resolving power would be great, but seems unrealistic and isn't strictly necessary. And again - consider the separation between the sensor and the strip(s) to be 0.5 - 1 m.

ChatGPT doubters - how much have you used it? I used it to write a 500-line python program that pulls data from an API and creates and animation. I basically just described what I wanted at a high level and it wrote the code. Not all in one go - I told it what I wanted step by step. But I hardly wrote a line of syntax at all. It gives good idea for which data structures to use, it decides its own variable names, and the code looked cleaner than anything I would have written. Try this - go find a short program or script and copy and paste it into GPT and ask "what does this code do?, how does it do it? and why might someone find this useful?"
Field of view,
comes down to basic physics,
The lower the frequency of the source, the bigger the focus system needed,
so a torch , light, has a small dish to get a narrow beam
a radar, radio, much longer wavelength, has a MUCH bigger dish to get a narrow beam,

That's one reason a laser or such light would be a great source, narrow

I'm assuming this is a university project , not a real product,

So you can do a good write up on angular resolution, and the reason radio is a bad choice.

https://jila.colorado.edu/~ajsh/cou... angular resolution is proportional,: q = l/D.

If you think of light for a second,
shine a big light at a black surface, with two bright spots on it,
you will get a "fixed" amount of refection off the black with the to spots
if one spot was covered the reflection would change
if both are covered, then a different reflection.

no need for a narrow beam, or receiver,

"just" some "simple" digital correlation / background averaging

your bypassing the angular resolution problem,

discuss..
 

Ya’akov

Joined Jan 27, 2019
9,070
It would be exceedingly helpful if you would explain the problem you are trying to solve. What do you hope this will do?

Frankly, I don't think your proposal is a practical one and you may well be reinventing the wheel. A description of the application will help us understand your actual goal instead of the way you've decided it might be solved. You've created a phantom problem by shifting to the solution before you know if there is some other way to go about what it is meant to do.

Some of the assumptions you've expressed and analogies you've used are wrong, so you are probably spiraling off into a non-starter "solution". Can you describe the problem this will solve instead of the solution?
 

Thread Starter

Corey Haddad

Joined Oct 14, 2015
22
Ya’akov - I am just curious if this approach could work. I am thinking this could be a way for a car to locate itself. And yes, I know there are 1001 other ways to do this. So suppose I had a pattern of metal strips in or on the pavement - I want a way to count them and to know their approximate spacing. Then it could serve as a crude barcode of sorts.

drjohnsmith - Ah yes, I remember now that if you put a mask in front of a telescope that has two holes cut in it which are on opposite sides of the main objective, the resolving power remains unchanged, while the light gathering power is (obviously) diminished. Same concept for synthetic aperture radar.

One thing about what you are saying - I am not going to be covering either reflector, just passing over them with the sensor.
Perhaps I could take advantage of specular reflections?
And are you suggesting two antennas spaced apart from each other?

Also - I am theoretically aware of the relationship between wavelength and resolving power, but have very little practical experience with designs.
 

drjohsmith

Joined Dec 13, 2021
852
Ya’akov - I am just curious if this approach could work. I am thinking this could be a way for a car to locate itself. And yes, I know there are 1001 other ways to do this. So suppose I had a pattern of metal strips in or on the pavement - I want a way to count them and to know their approximate spacing. Then it could serve as a crude barcode of sorts.

drjohnsmith - Ah yes, I remember now that if you put a mask in front of a telescope that has two holes cut in it which are on opposite sides of the main objective, the resolving power remains unchanged, while the light gathering power is (obviously) diminished. Same concept for synthetic aperture radar.

One thing about what you are saying - I am not going to be covering either reflector, just passing over them with the sensor.
Perhaps I could take advantage of specular reflections?
And are you suggesting two antennas spaced apart from each other?

Also - I am theoretically aware of the relationship between wavelength and resolving power, but have very little practical experience with designs.
If your in the UK, you will see on major roads white blobs on the road at various points, Thee are used by police cars for speedo calibration,

30 years ago, I worked in Bosch on a system that was patented, to guide cars in smart roads via RF, and markers in the road
including junction detection,

If your looking at this, you had better check for IP rights,
 
Top