Mobile Robot Communication Project

[Andyjm]

Hi all,

I'm toying with the idea of building some mobile gas detectors based on wheeled robots. The wheeled robot part is irrelevant right now, as I'm not even sure how well I can develop the basic concept! But here's the idea: let's say there's a mobile unit equipped with a gas detector that is set to alarm if gas levels reach a given value. This in itself isn't too hard (I knocked out a rough sketch for my Arduino last night with an MQ2 sensor I 'borrowed' from a mate). But I'd like to have more than one unit, each of which is capable of alarming all of the other units if its own alarm level is reached. So that's my question - if I disregard the mobility issue for now, what kind of issues am I likely to face as far as the interface between the units is concerned?

I hope I've explained the idea well enough! Thanks all.

 

[GreenGiant]

depends on how far apart the units will be, if they are within, say 20-30 feet then bluetooth may work, or zigbee. You could set up ultrasonic speakers and receivers and set it to look for certain "codes" so to speak, that would be a cascading effect.

The problem you will have with bluetooth and zigbee is range, and power (neither is super high power but constant polling and keeping in communication can be a bit of a drag) also zigbee is incredibly difficult to use.

Ultrasonic would be issues with interference, many things make noise that you dont realize.

 

[Andyjm]

Thanks for the reply.

I think Bluetooth may be the first thing to try, at least until I can prove the idea is feasible. Given the noise issues I think ultrasonic might be out, since I'd like to try the idea in an industry setting.

Thanks again for help, it's a good place to start.

 

[(*steve*)]

There are also some small 2.4GHz radios based on the NRF24L01+ that are much cheaper than bluetooth modules and are very clever. There are libraries that allow networks to be built up using these.

 

[donkey]

I wanted to add something here too. the design of your project is something that makes me wonder HOW you are going to do this.
will the robot work on a predefined course and the only communicating is the gas detector alarm? or is robot going to be controlled wirelessly as well?
if its the first than its super simple, if its the second I would definitely go with the NRF24L01+ chips.
the reason for this is they operate on the 2.4ghz frequency with over 100 channels to choose from so less likely to get interference and if you do then swap the channel. the second which is the part I love is they are transceivers. I hooked up 3 of these little beauties to get some extra range, the middle one just received the code then retransmitted it. it was the simplest booster I found.
also what sort of gases you looking at? depending on flammability choosing a motor will be fun. if you choose the wrong motor you will know when something flammable is around due to the sparks inside the motor making the flammable object combust.
I seriously wish you all the best with this project and would love to see designs or prototypes when you have them

 

[Andyjm]

Thanks for the advice!
To answer your questions - a predefined route for the robots was the way I intended to go, otherwise it's getting more complicated than I intended originally. I'm hoping that this way I can keep the coding for the mobility and the detection separate.
On motors, this was something I'd given thought to for the same reasons you mentioned. Having a gas sensing machine create a gas explosion kind of defeats the purpose! Unfortunately, even as a project if I intended the design to be used in industry then I need to consider what type to use, but I'd rather not get bogged down in ATEX, as I'm sure everyone has an idea how little fun that would be.
I'm new to the forums, but I see there is a section for posting updates on ongoing projects, so I'll post some stuff as I go.

 

[donkey]

ok now we are getting simple. next question for you is:
rather than having a robot move around an area would it not be better to install something like smoke detectors in these areas? that way rather than a robot being in the wrong place when a leak starts the detectors are stationery in the areas that need constant monitoring? I know its more expensive to have say 20 of these than 1 robot but lets look at the safety side.
if the area you monitor is laid out over 1000 square feet and the robot is on the opposite side of the complex, by the time it gets to where the gas is you may have a big issue.
I still say go with the NRF24L01's but you can actually use them to do a dual purpose role. each one is hooked up to the next they constantly transmit a signal saying all clear and their code. the base then identifies the codes coming through. when an issue arises lets say on sensor 15 it transmits warning and its code which the others all send back to base. you then know an issue has arisen in that area and the others are still transmitting ok. if the gas leaks out past 15 then the other sensors will show how far it has spread lets say to 14 and 16 because they will transmit warning and their code when gas is detected.
also if something goes wrong like an explosion god forbid the detectors will get damaged, this will mean they broadcast nothing. so if lets say 15 stops broadcasting you can send the fire brigade straight there.
the NRF as stated before has 100 channels. it can also send and receive 2mbps with (they say) 100metre range (I highly doubt that through walls and other interference).
they operate on 1.9volt to 3.6 volt range and are really good on power consumption,
these can also be hooked up to a raspberry pi if you want so you can make yourself a handheld receiver to carry around.
I seriously cannot explain how much I endorse these suckers and when you find them on ebay for $2 AU then you realise they are an experimenters delight.
you could actually use the detectors to pass on other info say to a patrolling robot.
the robot could goto the detectors location and use its sensors, if the sensor from the detector is different from the robots then it could send a signal saying the detector needs replacing or fixing. if it detects variants on 2-3 detectors then you know the robot needs fixing.
I hate long winded replies but this has given me some food for thought and here are my ideas to help.
I seriously hope it doesn't discourage you but instead makes you design a working prototype that covers every aspect of safety in which I figure you are trying to do

 

[Andyjm]

Again, thanks for the great help.

I think I need to go back to the drawing board for bit. I had hoped that I'd be able to put together something original and useful, but to be honest the more I think about it I think the premise might be flawed. The scenario you've suggested is basically what is is use at the moment, and if an idea doesn't improve it, well there's more thought required I think.

I'm playing around with some other ideas using remote communication right now so I'll still be taking the advice I've been given on that, so many thanks again for help!

 

[donkey]

sorry been occupied so haven't responded.
Andy I am of the firm belief your idea is a good one. I just see it as a backup. put it as a secondary or back up safety issue,
the robot is something that can also be used, or taken offline for repairs or put in one area if power outages occut.
do not make this the primary system but have it as a good back up and sell it to those promoting safety.