I just wanted to start a radar/warning system for a bicycle for fun. I know there are some on the market, that’s not the point.

I have already done an ultrasonic one using the same sensors as car reversing sensors and Arduino, but that has only a range of about 6 meters max, not enough range.

I have been experimenting with XYC-WB-DC 5.8GHz microwave radar motion sensor module (doppler). The following link shows the device I found on the net, https://www.e-gizmo.net/oc/kits documents/XYC-WB-DC Microwave Radar/xyc-wb-dc.pdf

I am not using the out pin shown in the PDF as that is only an on/off i.e. triggered pin that shows a movement was detected. I have tapped into the onboard amplifier and am getting the signal from the second opamp. So I get an analogue output voltage that represents what the microwave module is outputting when it detects movement. I feed that to A/D of Arduino and I can see the signal in the serial plot.

I would like to process this signal to get a rough idea of a moving target range. The idea is to place this module on the back of the bike and get a representation of an approaching car or other cyclists etc. my understanding of doppler type radar is that a target moving towards the radar creates a higher frequency and going away from the radar generates a lower frequency, I kind of can see that in the plot moving my hand towards and away from the unit.

I searched for FFT for Arduino but haven’t been able to implement anything that shows or results in anything useful yet. I am not looking to make a very accurate system, just something that gives a warning of approaching objects cars etc from behind a cyclist on a display of a sort or via BT on a mobile phone with audio warnings etc.

Currently, the module is very sensitive, I can detect movements through walls, so it should have a reasonable range in open space. I have not yet mounted the sensor/system on my bike to see how it reacts to being moved around whilst detecting.

So I appreciate any good ideas, especially on signal processing.

 

I have already done an ultrasonic one using the same sensors as car reversing sensors and Arduino, but that has only a range of about 6 meters max, not enough range.

According to the datasheet in the link, the radar module range is ~6-9 metres. Is that going to be enough?

 

According to the datasheet in the link, the radar module range is ~6-9 metres. Is that going to be enough?

It is omnidirectional though. It is possible that a corner reflector behind it would increase the range quite a bit.

 

@ Alec_t I will have to run some tests but if it does 9 meters, that would be better than flaky 6 meters of ultrasonic.

@Ya'akov You are right, I will be placing it in a shielded box and see how that works.

 

@ Alec_t I will have to run some tests but if it does 9 meters, that would be better than flaky 6 meters of ultrasonic.

@Ya'akov You are right, I will be placing it in a shielded box and see how that works.

Designing a proper corner reflects for 5.6GHz would be better, but maybe it would be enough to prevent detection from the back and sides, I don't know.

 

Yes, Ya'akov, I agree but for now, I just want to see if the module can be used for the intended purpose.

If the range of the one I am trying to use is not enough, this one seems to have a slightly longer range and seems to work the same way looking at the amplifier on its output.

 

Certainly Y is correct, and I would point out that a directional antenna can provide a lot of the least expensive gain while at the same time making it more directional. And already there exists a huge body of antenna information.

 

If mounted on a bike, how do you plan to subtract the movement of the bike itself?

 

If mounted on a bike, how do you plan to subtract the movement of the bike itself?

He only cares about the relative positions of moving things behind him, not their speeds, so it's irrelevant.

 

Certainly there will also be a fair amount of doppler signal return from stationary objects as they are passed, and so the DSP will need to be able to ignore that portion of the signal. So the signal processing will not be simple at all.
But even with a range of 30 feet a vehicle approaching at any serious rate will be a challenge to dodge. Or is this to detect other bicycles overtaking??

 

If mounted on a bike, how do you plan to subtract the movement of the bike itself?

I thought about that, I dont know yet, need to mount it on for test and record some data to see how it behaves and try to figure out what to do with that! I am trying to look at it in a simple way, not sure how much that would affect it when I am just looking for triggers based on reflection rather than a full-blown radar.

 

He only cares about the relative positions of moving things behind him, not their speeds, so it's irrelevant.

Exactly, trying to keep it simple, this is just to get an idea of roughly what is behind possibly 90 degrees.

 

Certainly there will also be a fair amount of doppler signal return from stationary objects as they are passed, and so the DSP will need to be able to ignore that portion of the signal. So the signal processing will not be simple at all.
But even with a range of 30 feet a vehicle approaching at any serious rate will be a challenge to dodge. Or is this to detect other bicycles overtaking??

This is a safety thing, so I am trying to get a warning out of this if anything is getting close. Of course, nothing beats looking around and being alert, but this is just another tool.
Yes, signal processing is not going to be simple but again I am trying to keep it as simple as possible, I havent done DSP before.
If you are going at 10 miles and a car approaches you at 20, I am guessing that would be a substantially different signal than stationary objects.

 

When I am going 15 MPH and a car 50 feet behind is going 45 MPH there are 2 seconds at best to achieve an evasive action.
So I suggest that in addition to the rate of closure range of 9 meters, a reflection detection range of closer to 100 feet is in order.
How about a rear-direction -listening device to warn of tire noise approaching. A parking distance radar does not seem to be worth the effort.. "Cool"? Certainly! but useful? probably not. UNLESS you need to be warned about others on bikes. THOSE can be hazardous.

 

When I am going 15 MPH and a car 50 feet behind is going 45 MPH there are 2 seconds at best to achieve an evasive action.
So I suggest that in addition to the rate of closure range of 9 meters, a reflection detection range of closer to 100 feet is in order.
How about a rear-direction -listening device to warn of tire noise approaching. A parking distance radar does not seem to be worth the effort.. "Cool"? Certainly! but useful? probably not. UNLESS you need to be warned about others on bikes. THOSE can be hazardous.

I get your point but, this is not about taking evasive action to save your life. This is about just a bit of warning/awareness of what may be behind you. In all commercial aircraft, there is a proximity warning system called TCAS , it warns of nearby aircraft in the air within a few miles of a plane. They dont have that just so that they can avoid other aircraft, its just another safety system.

My idea is to have something that aids awareness. this unit only works maybe 9 meters but there are other units that may be used for a longer range. Maybe this one is proof of concept, I dont know. For now, if it just warns of other cyclist behind, its still worth a play.
Listening for tire noise is the domain of DSP, not a job for me I am afraid, interesting though but then you have all sorts of surfaces and tires...
Here in London, probably most roads have become 20 mph roads so you wont have a 45 mph car chasing you. Also a lot of cycle lanes and detecting another cyclist approaching may just be useful.

 

In all commercial aircraft, there is a proximity warning system called TCAS , it warns of nearby aircraft in the air within a few miles of a plane.

Did you read how TCAS works? It's not really radar based.
" TCAS involves communication between all aircraft equipped with an appropriate transponder (provided the transponder is enabled and set up properly). Each TCAS-equipped aircraft interrogates all other aircraft in a determined range about their position (via the 1030 MHz radio frequency), and all other aircraft reply to other interrogations (via 1090 MHz). This interrogation-and-response cycle may occur several times per second "

https://en.wikipedia.org/wiki/Traffic_collision_avoidance_system

 

I get your point but, this is not about taking evasive action to save your life. This is about just a bit of warning/awareness of what may be behind you.

They have had things that do that since I was a boy. And that was a long time ago.
https://www.icebike.org/what-you-need-to-know-about-cycling-mirrors/

 

Did you read how TCAS works? It's not really radar based.
" TCAS involves communication between all aircraft equipped with an appropriate transponder (provided the transponder is enabled and set up properly). Each TCAS-equipped aircraft interrogates all other aircraft in a determined range about their position (via the 1030 MHz radio frequency), and all other aircraft reply to other interrogations (via 1090 MHz). This interrogation-and-response cycle may occur several times per second "

https://en.wikipedia.org/wiki/Traffic_collision_avoidance_system

Thanks, I know how that works, that was not the point. It was about having multiple warning systems. Ah yes, the mirror technology.

 

I get your point but, this is not about taking evasive action to save your life. This is about just a bit of warning/awareness of what may be behind you. In all commercial aircraft, there is a proximity warning system called TCAS , it warns of nearby aircraft in the air within a few miles of a plane. They dont have that just so that they can avoid other aircraft, its just another safety system.

My idea is to have something that aids awareness. this unit only works maybe 9 meters but there are other units that may be used for a longer range. Maybe this one is proof of concept, I dont know. For now, if it just warns of other cyclist behind, its still worth a play.
Listening for tire noise is the domain of DSP, not a job for me I am afraid, interesting though but then you have all sorts of surfaces and tires...
Here in London, probably most roads have become 20 mph roads so you wont have a 45 mph car chasing you. Also a lot of cycle lanes and detecting another cyclist approaching may just be useful.

OK, Been in London Town and I know that you are right. But I was thinking of the quiet lanes between Birmingham and Sollihul, that I explored during a weekend when not servicing a carburetor lab test machine for Austin Rover. That was quite a few years ago, about 1986 or 1987, I think.

 

OK, Been in London Town and I know that you are right. But I was thinking of the quiet lanes between Birmingham and Sollihul, that I explored during a weekend when not servicing a carburetor lab test machine for Austin Rover. That was quite a few years ago, about 1986 or 1987, I think.

You are a brave man MisterBill2, I see where the 45/60 mph car chasing comes from!

As a driver, I see how some drivers treat cyclists, and I don't want to start a discussion on that! I want to ride my newly acquired bicycle to work 3 miles away, but I am too nervous, being in central London I am looking to build any warning system I can to give me a little more confidence. I am a lot more confident flying a little light plane a few 1000 ft up than cycling in London!