IR sensors for an audio art installation

Hey

Noob here looking for advice on coding, hardware, and procedure for an interactive audio art installation.

I have a university master's project of connecting 50 IR sensors (digital) to an Arduino mega.
The IR sensors will be suspended from a squared grid net in a suspended ceiling (think bird wire - see link - Polyethylene (PE) netting | black | 80mm mesh size (HM) | twine no. T120 (1,2mm Ø) | width 16m | ENGEL-NETZE). The sensors will be attached to each cross-section and pointed down so that when people walk underneath them, the sensors trigger and send a signal to the mega, which sends to a Macbook Pro (M1) and then to Max 8 (Max/MSP). Each triggered sensor will be attached to a sound file within Max 8 and played out of a loudspeaker in the interactive area.

I have little to no coding experience other than a starter kit I tinkered with several years ago.
I have an Arduino Mega (R3 2560), also some IR sensors (not all 50 yet), plenty of wire, etc.

I believe I will need some kind of junction box that I can label and connect each set of pins to and also a power management system as each sensor will need 5v but not sure what to buy.

Any help will be much appreciated

Can you post a datasheet for the IR sensor please

Pretty sure this is the sensor ill be using

The max detection range is only 30cm (about 11 inches)

Ah, I see.

I had a feeling that might be the case, short-range, etc.
At this point in the project, anything can change, nothing is set in stone, so I'm completely open to any suggestions. For example, making the ceiling a wall instead so that participants can get close enough to trigger the sensors / manipulate sound, or even a different type of sensor that may be powerful enough to reach at least 3 feet.

Thanks for any information :smiley:

You might be able to use LASERs If the floor is white the LASER beam would be reflected back to a receiver. When someone walks under it, the beam would be interruped and you could detect presence

I found a VCNL4100 ( https://uk.farnell.com/vishay/vcnl4100/ambient-light-sensor-i2c-3-6v/dp/2627813 ) but 50 of them might be too much at £123. I'd have to reduce the number of sensors and it looks a little more complicated to get set up.

This HC-SR501 PIR ( Amazon.co.uk )
seems ok, not a laser but I think it might do the job, I'm just wondering though, as it is a human motion detector, its field of detection would be rather large and would trigger very quickly when a person walked into the interactive area, instead of triggering roughly when stood underneath a sensor.

You are right in both cases. The vcln is too complicated and a PIR will have a wide detection range

1 Like

Over what dimensions of floor will this be watching?

How many of the 49 grid points might be active at once, that is to ask how many audio files might need to play at the same time?

Where these audio files are stored, how many different ones and their maximum, minimum and average length might be nice to say.

Are you insisting IR sensors, or woukd the real problem be detecting the presence or absence of something in the area a grid point is monitoring?

a7

The area of interactivity will be a 7'x7'x7' (2.13x2.13x2.13m) truss rig setup (see image)

I would like only one person at a time to use the interactive space, so the number of activated sensors could be around 15. This is just a guess but it may depend on the size of the person and if they run through it or spin, etc. (I may have to advise people to use the space slowly?)

Using Cycling74 - Max 8 software, I am hoping to map all 50 IR sensors to individual sounds, each will be quite short like a chime or a piano key length, so let us say 3 seconds each.

For the type of sensor, I would like to keep it as cheap as possible, IR sensors are around £25 for 50 of them. If there is a cheap alternative then I'm totally open to trying it depending on complexity and cost.

Webcams have been suggested or an Xbox 360 camera but it seemed too complex and has been previously attempted, therefore, I am trying to use ideas that are alternative to the university's previous student's work.


I think people are going to suggest video camera and image processing, which is outside my experience but seems like a thing, easy and inexpensive.

Ooh, never mind, actually read what you wrote.

If that was very long ago, check again. Progress gets made…

a7

You might consider the Melexis MLX90640 thermal imaging camera, as it is an all-in-one solution to detect a person as well as determine the person's location. Two focal lengths are available.

1 Like

This will work if ONLY one person is in the area

This is the cheapest I could find

1 Like

Wait it will work with everal people

Ok, thanks. This is all great info.

For this project, I'm trying to keep the costs as low as possible ( student here :laughing: ), I love the ideas of all the tech so far but maybe low cost means low proximity sensors, etc.

Let's say I wanted to go ahead with my design (shown in the above post).
Would I need a separate power supply for all the sensors?
What kind of connection/junction box would I need (if any)?
Despite the 30cm proximity of the sensor, is this project doable?

You can determine that starting with one sensor. If one doesn't work in the project environment, 50 won't.

Yes, you will need a separate 5V supply at least 2A.
You should use18 gauge wire to connecting everything. You can use wire nuts or euro strips for connections.
Yes it is dooable (if you get within 1foot)

1 Like

even better

1 Like

Perfect, thanks jim-p.

I will look into these parts