Pondering a movement activated / controlled fence line

[scamper]

Hi guys. I have been out of the xmas lighting for a few years as we bought a new business a couple years ago that has taken ALL of our time. Finally I can see light at the end of the tunnel as we have managed to almost finish moving premises and getting some employees to take the pressure off.

Sooo......
We have a fence frontage that is approx. 20m long and is about 600mm above ground level with a flat grassed front verge. We often have kids (and parents) dancing on the verge to the music and I wanted to have some kind of movement control of the fence line where as someone moves up and down the fence line the lights will "follow" them.

I was thinking ir sensors, but they, I would assume, would be too hard to narrow a beam and would just go off almost randomly, Then I thought some kind of infrared beam, but a reflector would make that impossible.
I am not 100% sure on ultrasonic but think I may do some testing with them and an Arduino
Has anybody had any experience using these that they can offer or indeed has anyone done anything like this that they offer their advice?

 

[scamper]

So before anyone says anything about the sensor I pictured, it is only a test as I have them in my kit. They are not waterproof and only have a range of about 500mm
I would end up using something more appropriate with waterproofing and a better range. But same applies if anyone has experience, advice always welcome.

 

[TerryK]

The ultrasonics I've worked with (in industry) were a bit prone to false triggering which with pixels might not be a problem. Units too close together also caused problems. Sensing field width relative to distance was another concern. I would as you mentioned set a couple up on a bench and see what separation they need and how good is their detection distance. Not sure about the number you are thinking about (detection resolution) but rather than utilizing 20 or more inputs perhaps research a method to multiplex them.

There are beam type IR sensors with focused narrow beams. You ruled out the retro-reflective sensors due to needing a reflector but keep in mind that a reflector isn't always necessary if there is a 'beneficial' background. You might also look at diffused IR sensors. Problem with those though is colors that the device's beam would reflect from; skin/clothing. I detested working with these but sometimes they were a good fit. There are also the emitter/receiver IR sensors. Might it be possible to have a strip light of sorts illuminating the dance zone? It would need to be placed past the detection zone though. The light would simulate/replace the emitter. The light would not need to be a beam, a fan or line style can be made to work reasonably at times. Receiver sensors would or could be dotted along the fence. Another point to consider is IR sensors with a wide detection field can have its field narrowed by making it look through a cylinder.

 

[Kent]

Starting to get a little more expensive, but Seeed Studio have a number of lidar based options from < $100 that are IP65 rated, e.g:

TFmini-i LiDAR - Distance Sensor (12m) with CAN Interface

TFmini-i is a single point LiDAR, which is based on Pulse Time of Flight (PTOF). It adopts an incoherent energy receiving mode, and the detection is mainly based on Pulse counting. TFmini-i emits a narrow pulse laser, which is collimated by the transmitting lens to form a collimated light, which...

www.seeedstudio.com

That one above talks about 12m range, but some of the more expensive modules go up to around 180m range:

TF03-100 LiDAR – Industrial-Grade Long Range Distance Sensor (100m)

TF03-100 LiDAR is an industrial-grade long-range distance sensor. Its maximum detection range can reach up to 100m and it has an adjustable frame rate with a maximum of 1KHz. With an integrated compensating algorithm for outdoor glare and other interference, TF03-100 can work under a strong...

www.seeedstudio.com

 

[merryoncherry]

It's probably too low tech for this application but I use PIR sensor to trigger the "light up the outline and turn on the floods" sequence for selfie spots.
If you really want to follow them, you might need a camera... or I remember when everyone's project had to use a Kinect.

 

[scamper]

It's probably too low tech for this application but I use PIR sensor to trigger the "light up the outline and turn on the floods" sequence for selfie spots.
If you really want to follow them, you might need a camera... or I remember when everyone's project had to use a Kinect.

Hmm. Camera
I remember back in the 80's working on a movement activated sensor based on a B&W camera picture. All it did was compared the average light output and as that changed it would trigger a switch. I am not sure my old grey matter is up with designing that kind of circuitry these days, I got out of the electronics industry 20 years ago and this hobby is one of the few things that I even consider now.

 

[David_AVD]

You can narrow the field of view of the small PIR sensors by recessing them just a little in a tube. That's what I use for some of my Halloween props.

 

[Notenoughlights]

Wonder if you could use the motion tracking sort of equipment they use for VR setups, a couple of CCTV cameras and some code to control the sequence?

 

[David_AVD]

Here's some more info on the PIR I use and a mod to decrease their ON time to make it more responsive.

HC-SR505 PIR – da-Share

www.da-share.com

 

[scamper]

Here's some more info on the PIR I use and a mod to decrease their ON time to make it more responsive.

HC-SR505 PIR – da-Share

www.da-share.com

That looks more my price range
What sort of range do you get with these? Ideally I would like a couple of meters which would cover the verge, Maybe if it were a few more I would surprise a few cars that drive by on quiet nights

Also, the "on time" is that the time it stays on? or delay in turning on?

Depending how much time I have I may still experiment with ultrasonics as I like the idea of being able to not only control the lights going on and off, but am thinking that the distance away may be able to control the colour of them too

 

[David_AVD]

They detect pretty much instantly. The ON time is how long it extends the output active time.

I think you'll get at least a metre from them, possible 1.5m - 2m.