speed sensor needed

[peanut]

Hi All
I need to measure the speed of delivery vehicles down a private lane .
The speed range will be 5-30mph
Can anyone suggest a readymade solution or an easy diy solution?
The lane is approx 80 metres longI can arrange detectors along a 8 metre stretch .
Any ideas welcome
peanut

 

[Nigel Goodwin]

Hi All
I need to measure the speed of delivery vehicles down a private lane .
The speed range will be 5-30mph
Can anyone suggest a readymade solution or an easy diy solution?
The lane is approx 80 metres longI can arrange detectors along a 8 metre stretch .
Any ideas welcome
peanut

The simpliest way would be to have two IR beams, and measure the time between them breaking - you could do the timing with either hardware counters, or a PIC could do it with a single chip.

You can also get air-pressure operated switches, you could lay two hose pipes across the road, both feeding to a pressure switch, and measure the time taken between switch activations in the same way.

 

[peanut]

Hi
thanks for the quick response.
I have googled for hours and found nothing yet?

I want this to be undetected if posible so I would prefer to use a couple of discreet IR units.
How would I store the trigger times and convert the time betweentriggers into speed?
I would like to be able to trigger a camera if possible to record the vehicle

 

[Nigel Goodwin]

Hi
thanks for the quick response.
I have googled for hours and found nothing yet?

I want this to be undetected if posible so I would prefer to use a couple of discreet IR units.
How would I store the trigger times and convert the time betweentriggers into speed?
I would like to be able to trigger a camera if possible to record the vehicle

As I mentioned, the easiest way is to use a PIC (or other microcontroller), simply wait for the first beam to be triggered and then start counting, as soon as the second beam is triggered you stop counting. The value you have counted up to is an indication of the speed, the higher the number the lower the speed. You can then calculate the speed based on this count, obvious variables are how fast you count, and how far apart the beams are - you can easily compensate in your software for these, the counting speed being the obvious way to calibrate it.

Another advantage of a PIC solution is that you can store the results and display them later, or trigger anything you want from an output. It also akes it easy to add extra functionality, such as only videoing vehicles going over a certain speed.

As for your googling, I would suggest you try for model cars - I'm sure I've seen similar speed units for slot car racers etc.

 

[peanut]

excellent thanks for your suggestions.
I am unable to program PICS myself although I used to know someone who could.
I would like to build something myself but I need an urgent solution before there is another fatality down my lane.
I need to be able to produce proof so I would prefer to buy a readymade solution if poss.
Nick

 

[zevon8]

One thing that may not only give you proof, but be a great deterent, is to install a video camera, and advertise its presence with a sign. Most people are aware of video technology, and when they are breaking the rules they don't like witnesses...

I have seen similar set-ups in parking garages, the sign reads something like " Speed Limit 10 Mph -- Video Surveillance in Use". Even if there are no real cameras, most will think twice.

 

[Oznog]

Outdoors, you will need a AC (pulsed) signal because ambient infrared in the daytime will probably be 100x brighter at the receiver than your transmitted signal- but sunlight does not change quickly thus can be effectively filtered out.

Since there are two transmitter/receivers operating in close proximity, you may need to give each one a different pulse train and guarantee that they do not transmit at the same time.

A more focused receiver and/or transmitter will also considerably improve your signal-to-sunlight ratio. This may also fix the problem of the receiver seeing the signal from the wrong transmitter.\

And yes, a microcontroller (PIC) is the only really good way to do this.

 

[peanut]

One thing that may not only give you proof, but be a great deterent, is to install a video camera, and advertise its presence with a sign. quote]

Iwould like to do this but the trouble is its a single track private access road to 7 houses owned by the residents and my neighbours do not want the site `spoilt' by having signs errected like a `parking lot.'

The other problem is that some of the neighbours are actually guilty of speeding themselves (especially the women) but all adamantly deny it when challenged. They claim they are driving at 5mph but are actually doing 20-30mph!

If I could show them actual visual evidence in one of our residents meetings then I might be able to shame my neigbours into slowing down and also agreeing to errecting perhaps one sign.

To do this I need to be able to show them a video clip together with a speed reading so that it is beyond dispute

 

[peanut]

Outdoors, you will need a AC (pulsed) signal because ambient infrared in the daytime will probably be 100x brighter at the receiver than your transmitted signal- but sunlight does not change quickly thus can be effectively filtered out.

Since there are two transmitter/receivers operating in close proximity, you may need to give each one a different pulse train and guarantee that they do not transmit at the same time.

A more focused receiver and/or transmitter will also considerably improve your signal-to-sunlight ratio. This may also fix the problem of the receiver seeing the signal from the wrong transmitter.\

And yes, a microcontroller (PIC) is the only really good way to do this.

I can obtain IR tx rx stuff very cheaply from Maplins or RS etc but I am going to need a circuit board design for a microprocessor controlled solution.
I assume the IR can be made fairly directional and could be arranged opposite each other across the width of the road so that the beam is broken by the cars. The 2 stations can be sited at least 7 metres apart which should prevent any cross triggering.

Presumably I need a circuit that will register the time delay between triggering point A and B convert that into a speed and display that on a led also trigger a video camera which has the led display in the view window.

 

[Nigel Goodwin]

To do this I need to be able to show them a video clip together with a speed reading so that it is beyond dispute

In which case it would be helpful to record the speed reading on the video tape, you can buy PIC controlled onscreen display modules, which would allow you to superimpose the reading on the video signal.

 

[peanut]

Thanks for all your suggestions . Sorry for the slow responses but for some reason I am not being updated of your replies although I have subscribed to the thread. I'll check my settings.
The next question is where can I go to get someone to design build and test this for me ?

 

[stevez]

If evidence of speeding is what you need then the video camera/recorder, correctly placed, may be enough. I know security type video recorders are such that you can advance one frame at a time. If you could do a playback at one frame at a time you might be able to crudely calibrate an arrangement. That would require that someone review the tapes - which might be tedious.

I wonder if the cameras that can be used on computers could be used - with the precise time down to split seconds recorded on each frame. The thought here was that you might already have a computer and might borrow the camera from a friend - and maybe the software is low enough in cost.

 

[peanut]

ingenious idea. I have loads of shutle pcs sat doing nothing I could buy a 4-8 camera capture/ input card with security software. I had been considering doing this for some time. You can set the system to record onto hard drive automatically and view the property live through a BB net connection from anywhere in the world.

Trouble is I know of no way to calibrate picture frame rate / distance treaveled into a speed readout ?
It would be perfect if it could be done as I could superimpose the speed in real time on the video clip.

I could set one camera to look up the lane and one set to look down. ( I'm in the middle) Then set two directly across the lane with a measured gap between. mmmmmmm

 

[stevez]

5 mph is about 3.5 ft/sec. If the camera were placed so that is spanned maybe 50 ft of roadway you could simply play the stuff back at normal speed and use a stopwatch. If the vehicle takes less than 14 seconds to travel 50 ft it's going faster than 5 mph. If it's 7 seconds it's doing 10 mph. I don't think you need a high degree of precision here - just order of magnitude.

The camera must have a frame rate or pseudo shutter speed - it's probably well known - just that we don't know what it is. If your software can allow you to advance a frame at a time then review would not be too tough. A lot of security camera stuff timestamps the resulting video and if it goes out to enough places you'd be all set. Distance/time is speed.

If you had a motion sensor trigger the 'record' feature for a limited amount of time you'd eventually be able to easily judge who is going too fast. Example: camera is on for 7 seconds with field of view that is 50 ft of roadway. If the vehicle leaves the field of view before the 7 second segment is done it's going more than 10 mph. If it makes it halfway across it's doing 5 mph. That might help to limit the amount of video stored and gives you enough resolution for what you are doing.

Even if your camera is at an odd angle you could 'calibrate' it by driving a vehicle at known speeds then comparing the time it stays within the field of view or the time it takes to pass certain markers that are in the field of view. Again, not precise but probably good enough.

 

[peanut]

there is some good ideas here thanks but there are a number of problems however.
Firstly in order to confront someone about their allegeded speeding or dangerous driving I will need to be able to establish their speed beyond reasonable doubt and I will therefore need a bit of software or microprocessor which can be calibrated .

It will not be enough to say " your car passed this point in this many seconds therefore by deduction and observation I estimate you must have been driving at x mph" they will simply deny it and I will get nowhere.

The problem is most of these people are habitual speeders and genuinely don't realise/accept they are driving at 20-25mph down a 9 foot wide private track past peoples doors to their houses.!( doorways are within 3 feet of the road)

I need a bit of software macro or a programmed PIC or Stamp etc
I imagine there must be something commercially produced to measure speed of things like conveyor belts etc ?

 

[Roff]

5 mph is about 3.5 ft/sec. If the camera were placed so that is spanned maybe 50 ft of roadway you could simply play the stuff back at normal speed and use a stopwatch. If the vehicle takes less than 14 seconds to travel 50 ft it's going faster than 5 mph. If it's 7 seconds it's doing 10 mph. I don't think you need a high degree of precision here - just order of magnitude.

The camera must have a frame rate or pseudo shutter speed - it's probably well known - just that we don't know what it is. If your software can allow you to advance a frame at a time then review would not be too tough. A lot of security camera stuff timestamps the resulting video and if it goes out to enough places you'd be all set. Distance/time is speed.

If you had a motion sensor trigger the 'record' feature for a limited amount of time you'd eventually be able to easily judge who is going too fast. Example: camera is on for 7 seconds with field of view that is 50 ft of roadway. If the vehicle leaves the field of view before the 7 second segment is done it's going more than 10 mph. If it makes it halfway across it's doing 5 mph. That might help to limit the amount of video stored and gives you enough resolution for what you are doing.

Even if your camera is at an odd angle you could 'calibrate' it by driving a vehicle at known speeds then comparing the time it stays within the field of view or the time it takes to pass certain markers that are in the field of view. Again, not precise but probably good enough.

5 mph=7.33 ft/sec

 

[stevez]

Ron - I stand corrected- 60 mph is 88 ft/sec - probably inverted the ratio. Oh well, concept or idea is still illustrates that the idea might work - distance would have to double, approximately to use time periods I described.

 

[Roff]

Ron - I stand corrected- 60 mph is 88 ft/sec - probably inverted the ratio. Oh well, concept or idea is still illustrates that the idea might work - distance would have to double, approximately to use time periods I described.

I suspect that a third of my posts here and on other forums are "nit-picks" - I just hate to see hobbyists get led down the garden path. Not that I haven't done it myself a time or two. :cry:
Maybe I shoulda been a proofreader instead of an engineer.

 

[stevez]

Ron - not sure about you but I post here in between phone calls, conversations or while waiting for other stuff. I suspect others do the same. Errors are easy to make - but the important thing for me is to illustrate the thinking process, do the math, so to speak, which helps the originator and other readers to carry on from there. Corrections are never a problem, in my mind. As you say, best to make sure the right information is out there. Now and then I'll see errors that jump right off the page and I point them out. I don't worry too much about being a little wrong nor pointing out errors. That's all a part of the interaction that makes it a good forum.

 

[Roff]

Ron - not sure about you but I post here in between phone calls, conversations or while waiting for other stuff. I suspect others do the same. Errors are easy to make - but the important thing for me is to illustrate the thinking process, do the math, so to speak, which helps the originator and other readers to carry on from there. Corrections are never a problem, in my mind. As you say, best to make sure the right information is out there. Now and then I'll see errors that jump right off the page and I point them out. I don't worry too much about being a little wrong nor pointing out errors. That's all a part of the interaction that makes it a good forum.

I agree, although I really stuck my foot in my mouth here one time - right, Nigel? I still get thinking about it.
I've been involved in a friendly (?) argument on the N&V forum during the last few days. I got involved because I could see that the proposed solution was going to cause the OP to burn up some power transistors if he followed the advice being given. I jump into those reluctantly because N&V, like this forum, is polite and civil (to a fault, sometimes), and criticism sorta looks like flaming, but I hate to see a hobbyist smoke parts that he may not be able to easily replace. That obviously was not going to be the case here - maybe just a little embarrassment on Peanut's part if he didn't catch the error.