nurse call system

hi all,

I have a project were I have to create a nurse call system, for approximately 50 beds.
Basically each bed will have a button, and in reception a number will be displayed accordingly.

I dont know what is the best way to go around it. Shall I make a micro arduino in every switch and send different pulses to the main arduino board?

or, make a different resistance value in every bed switch? but here i am concerned about the cable resistance, which may alter the results.

or there is an arduino attachment board, that is ready available?

thanks.

How about a Mega, although it depends on your definition of "approximately 50".

Run a pair of wires to each bed from a pin with pullup enabled. Poll the pins in loop, if pin goes low (=button pressed), display pin number on LCD. (Or use pin number to index an array of bed numbers.)

(It's a strange hospital that doesn't have this already though, all wired in nicely in the trunking on the wall with the tv wiring and the other stuff.)

That sound like an interesting project :slight_smile:
I work in telecoms which means I am used to running cables around buildings. That part, running the cables, is going to be a big part of your project, much more than the electronics.

My thoughts are:
You call units need to be as simple as possible, as you will have 50 of them. The complexity needs to be in the central unit, so no to putting an Arduino in each one.

You could have a pair of wires to each call unit and have the central Arduino scan the pairs for a button press. There are loads of examples on this forum of how to read large numbers of button inputs. Advantage of this method is that a problem (short, disconnected wire, water spilt into the call unit etc) will only affect 1 call unit, not several of them. Disadvantage is the amount of cable needed and having to wire up 50 pairs of wire at the main unit.

You could have a loop, or possibly several loops, with the call units being switches that have resistors across them and are normally closed. Operate the call unit and the switch opens, putting the resistor in series with the loop. You use the A2D to measure the resistance of the loop. 8 circuits like that each with up to 8 call units will do what you need. You don't need to be too concerned about the cable resistance, it will be a few ohms to a few tens of ohms, depending on the cable length. You can use resistors of at least 10 times the expected maximum cable resistance.

You will need to consider the electrical noise picked up by the cable and filter it out at the Arduino. Effects can range from incorrect readings, false triggering, through to damage to inputs from voltage spikes.

Please keep us updated on progress, this sounds like a really interesting project.

You'll also need to be sure that whatever you do can be signed off by whoever has the authority to say it's not going to mess with all the gazillion of equipment. (I'm not saying it will; you may however have to prove it won't.)

I can not imagen this is some real situation but who knows.

If it is or not, might there be more information needed from every bed/patient in the future?

You need to look carefully at failure modes too - you might want to use change over switches at the beds and prove the switch is in either state , or run seperate check cables .

A lot of commercial system from my Googling are going wireless - as mentioned wiring cost is high and wireless can be more flexible . There are some standards with which to comply , and you’ll need good indemnity insurance and backup for when it goes wrong.

As others have implied , not the easy project it first seems , not really a hobby computer project .

You don’t want someone dying because your software has crashed and the lawer to find you adapted a hobby computer and took advice from non professionals ( eg me) either , not good that.

Please state if this is a school project or for use with humans
The Aruduino is not for use with any life safety application.

for a beginner, using a MEGA would allow you to bring in one wire for each switch. it would be much simpler for someone new to writing code.

for someone a bit more advanced, an UNO and some shift registers or a port expander would allow you to bring in a lot more switches with only a few pins on the board.

moving from one Arduino for the entire project to one at each switch :

also for someone a bit more advanced, using a tiny micro and an RS-485 at each switch would allow you to use one network cable. the balance goes to more expensive switches, but less expensive wire.

[ anyone know if a CAN bus might work ? node limitation ? ]

as noted, wireless is a possibility, however, the application might not permit radio use.

You will want this system to continue operating when there is a power failure.

Paul

I developed a POCSAG based wireless system few years ago... these units also transmitted the pendant battery level as a kind of closed-loop monitoring the pendant being still ‘alive’.

In a wired system, I’d suggest opto-isolated wiring runs - and N/C wiring to ensure the wiring and buttons are in place. (btoken wire or pressed button indicates a fault or call)

Of course, your choice to run 50 wires for 50 rooms, but i’d be considering a bussed system of some type o reduce the amount of wiring - and perhaps some extra features - like annunciator outside each room/area being wired.

hammy:
A lot of commercial system from my Googling are going wireless - as mentioned wiring cost is high and wireless can be more flexible .

Not in hospitals as it creates many problems.
Nurse call systems I've had anything to do with are on a bus system.

Just aside -
Every wireless pendant call system I've seen (used in aged care and nursing homes) is POCSAG radio based.
It seems that standard has been blanket ratified (not sure)
Note the ref in Australia - may be different elsewhere

hi all, sorry for late reply as I was busy trenching walls, and also did not have notifications on.
So, to answer a few questions, being the most important one, is that No, this project wont be used in a hospital, and it wont be used in life or death situations, so relax. :slight_smile:

As regards to cabling, yes.. I know it requires a lot of cabling, but being in cctv, intercoms, telecoms and networks, this cabling is next to nothing, also we are trenching for all power points, tv, tel, lighting, fire, burglar, etc in this building, so im pretty comfy there.

Back to the project, today i thought of using an AT mega, since it gives me plenty of inputs. As regards to the voltage drop, I thought of using these 16 channel relay boards. Each button will trigger a different relay, and each relay output will have a different resistance. This way I am eliminating the possible voltage drop.
Of course, you can tell me that its easier to use the board as latching, and build a simple panel with 50 leds.. and you will be right!. (I built same system with solid state relays 20 years ago, and its still going), but I just wanted to have it digital displayed... in 2019. :slight_smile:

What do you think?

Thanks for all replies.

elvon_blunden:
How about a Mega, although it depends on your definition of "approximately 50".

Run a pair of wires to each bed from a pin with pullup enabled. Poll the pins in loop, if pin goes low (=button pressed), display pin number on LCD. (Or use pin number to index an array of bed numbers.)

(It's a strange hospital that doesn't have this already though, all wired in nicely in the trunking on the wall with the tv wiring and the other stuff.)

its not a hospital

intertronic:
its not a hospital

Where do you keep the nurses you are going to call?

Paul

Paul_KD7HB:
Where do you keep the nurses you are going to call?

Paul

When I was 17, the Local General hospital matron used to keep all her nurses on the second floor level.
Hell of a climb on Saturday night after a few stubbies.