next topic »

[on:]

Hello everyone,

My name is Manuel and I'm from Copenhagen, Denmark.
I've been asked by various municipalities to develop a system which I am not to talk about yet unfortunately.
What I can say is this:

I need a "device" with which (either include or with which i can interface a IR sensor) I can perform measurements (distance) ranging 20 to 150cm top.
This is done outdoor, but inside an object.
This device will have to be able to send data to a centralized computer (or web server) running in the city hall.

I thought about 2 scenarios:

Scenario 1)
We take a small municipality area, which will be divided into sub areas.
Each of these sub areas will have MANY of these 'devices' (which have a long range WiFi ) and probably some routers once in a while for each subarea in order to more or less map the all municipal area creating so a kinda virtual network (or is it just a straight LAN? ) (Municipalities in Denmark are not very big--> 5 1/2 milion people live here, no mountains, not crazy buildings etc so not much interference or problems that these could cause).

So given the setup we will probably be able to just use WiFi to send information to a centralized computer in the city hall. (can the data that the sub-area router gathers, be forwarded to another router and forwrded again to the centralized computer? i would think so).

Plus:
-Everything would go on WiFi, easier and fast sending information and we wouldn't have a whole lot of information possibly bombarding our server.

Minus:
-More expensive (MAYBE???) than scenario 2, due to Routers costs
-Introduce a single point of failure, so in case the router has any sort of problem all the devices depending on it won't be able to send info to the centralized server.

Scenario 2)
Have a device into each object without having routers around, and use GPRS to send data (can I send this data to an ip corresponding to a webserver running in the city hall?)
Can someone please briefly explain me how sending data over GPRS would work? Do I need to install a Sim card on each waspmote and therefore paying the traffic data this way, or how?!?
I'm sorry but i lack of knowledge in this matter.

Plus: Maybe cheaper??

Minus: Might be more hard sending data to the server.

I would GREATLY appreciate if you guys could help me out on this, I really need.

[Reply #1 on:]

I've been asked by various municipalities to develop a system which I am not to talk about yet unfortunately.

pity, real requirements can be interesting...

I need a "device" with which (either include or with which i can interface a IR sensor) I can perform measurements (distance) ranging 20 to 150cm top.

you could google for the SHARP 2Y0A02 F 9Y  - but I'm not allowed to say that yet

Some questions arise:

How far are the sensors apart?
How often do they need to sample?
What is the expected precision?
What is the expected accuracy?
Should a timestamp be included?
What if a sample is wrong?
WHat if a sample is missed?
WHat powersupply do you intend to use?
Do you need watertight casing?
Do you need other sensordata as well? (often there is a need but people don't say that ...)
What are the operating conditions? Temperature, Humidity etc.
(be aware that the speed of sound is altered by temperature and humidity in a non-linear way which mangles your distance readings ...
Did I allready say you need extra sensors?..
....

In short, so many questions arise that it makes it almost impossible to give you any sensible advice..

Hopes these questions help you to focus on the technical difficulties,

[Reply #2 on:]

Hey thanks again!
Let me try to answer those questions:

How far are the sensors apart?
it varies pretty much. We could go from 35-40 meters to many km.
How often do they need to sample?
This doesn't have to be done very often: I'd say every 5 to 10 minutes.
What is the expected precision?
I wouldn't know of to "measure" this, put it into words. It has to be decent, the more consisten, the better
What is the expected accuracy?
This is not an issue.. I could tollerate a -+ 1 cm.
Should a timestamp be included?
A timestamp 'could' be included but it's not a requirements.
What if a sample is wrong?
If a sample is wrong is just discarded, no biggie. we'll examine the next sample.
WHat if a sample is missed?
Same as above, no big deal.
WHat powersupply do you intend to use?
Power supply i have no preference. A battery of some sort, but it does not really mater. Although it might be cool for future add on, a possibility to use a small solar panel to recharge it.
Do you need watertight casing?
well, this could be useful indeed! I knew we'd have needed a case of some sort.
Do you need other sensordata as well? (often there is a need but people don't say that ...)
Well, it'd be very very useful to a a gps in order to provide our location when sending the data.
What are the operating conditions? Temperature, Humidity etc.
This will be outdoor but inside an involucre.so it'll be shielded from all (or most) of the atmospheric agents. But just in case it'd be nice a watertight case (if this doesn't cost crazy)
(be aware that the speed of sound is altered by temperature and humidity in a non-linear way which mangles your distance readings ... --->> Is this the case of infrared as well? I'd have thought only ultrasonic sensor would be afflicted by this.
Did I allready say you need extra sensors?..
Yes you did!

THank you so much for the moment!

Manuel

[Reply #3 on:]

That is a very impressive project. Not for a noob, not a particular person either. And it not going to be cheap either.

I just want to add this :

IF someone work for the city of Toronto, and the GTA municipalities ask that person to do your type of project, the city and the GTA cities would hire a engineering firm - by placing an ad for the others firms to place a bid, a firm would be selected to do the project.  And the money involved are in the thousands and maybe millions here of tax payers money.

Example : Red light camera system all across Toronto. <-- My guess, the cost for that system is maybe in the millions. I would be surprise if the total cost is in the thousands.  

Edit - add :

But what you can do, is to design a prototype - one unit only, to test, to measure and to try and to install at one test location : Example : a very busy intersection.

And my guess of the system you are talking about :  A people monitor - counting peoples ? red light control ? animals and birds control ? crime control ? 

[Reply #4 on:]

That is a very impressive project. Not for a noob, not a particular person either. And it not going to be cheap either.

I just want to add this :

IF someone work for the city of Toronto, and the GTA municipalities ask that person to do your type of project, the city and the GTA cities would hire a engineering firm - by placing an ad for the others firms to place a bid, a firm would be selected to do the project.  And the money involved are in the thousands and maybe millions here of tax payers money.

Example : Red light camera system all across Toronto. <-- My guess, the cost for that system is maybe in the millions. I would be surprise if the total cost is in the thousands.  

Edit - add :

But what you can do, is to design a prototype - one unit only, to test, to measure and to try and to install at one test location : Example : a very busy intersection.

And my guess of the system you are talking about :  A people monitor - counting peoples ? red light control ? animals and birds control ? crime control ? 

Hey thanks for the heads up.
Basically my father in law has had an idea with a friend of his. And various municipalities has liked the idea.
Therefore he asked me if I could work on this, but I am more of an IT engineer (I'm about to graduate) and I could use this as bachelor project.
I can do the software application which we would use to monitor this data, no problem, but the whole process is more challenging.
I'm going to have a more experienced engineer friend of mine to help me out occasionally with the whole process and have other 2 guys graduating with me, helping me (we'll do our bachelor together).

You were close in guessing but i believe it's easier than those.Look at it if those measurements are going to be performed inside a box placed somewhere outdoor.

And you were right about the prototype!
What we are going to start on now is making a sample. We are going to try to implement a couple of these items and see how it will turn out.
We're going to calculate the costs, make some sort of financial prospect for the municipality to see the short term and long term advantages they would have, both economically and environmentally and we'll take it from there.

Of course if this were to be a success than we will have to rely on a company (I've already a couple of contacts) for the devices etc.

What I found to be the kind of device that would probably be the best fit is these:

http://www.sensor-networks.org/index.php?page=0932832814

http://www.libelium.com/products/meshlium/wireless-sensor-networks

Do you know them?

[Reply #5 on:]

Be aware that there are country specific regulations that in many cases could limit the power output of any wireless network node. Therefore the ranges quoted by these sensors will not be achievable.
I am not aware of the specific requirement in Denmark but it is worth checking them out before you spend any money.

[Reply #6 on:]

Yes that is true! Thanks, of course I will submit all the legal documents this company provides to the municipality or directly to the ministry of Science and Innovation.

Anyway I read the legal documents provided by the company and it appears that there are no restrictions for denmark nor license required (unlike in Germany, Italy, Norway etc), but for sure I will ask!

Anyway no one has any idea if (and where) i could find a device as such in order to have an alternative source i could evaluate my decision upon?

Manuel

[Reply #7 on:]

Using wireless sensor system network, when data are importants ? <-- I think, on a municipality property ( street light - traffic light - intersection corner ) can easilly be hack.  My idea is hard wires using ethernet system ( via the internet ) and all data got to a central computer to analyse and display the data.  When a system is hard wire, the box contain the sensor can be seal and lock so nobody can temper with or access with, unless you physicaly brake the lock. So you have a lot of tiny computer ( microcontroller ) connect to the net giving data to a central location.  Sorry , no wireless. If it is wireless, someone park can acess the data with a laptop and a wireless network card. With the right skills, it can be done. That why I prefer hardwire, just like the phone line.

As to design, build a prototype at your home, install on your property, data send to your home computer with a software design to received and process the data.  If every work, contact a engineering company who has a contract with the municipalities to impliment this type of system. Beside, when you deal with the governement ( local, state / provincial / federal ) , it have a lot of "red tape" to go through. I don't know about Copenhagen, Denmark... but this is the way when you deal with the canadian system... ( up to my knowledge, anyway )   

[Reply #8 on:]

We were testing something similar when I was in Asia. We basically created a wireless network using Ubiquiti equipment, then used the Ethernet shield to get the arduinos online at different points(the sensors where monitoring pumping stations with about 1km distance from the control center). We did not get to test the system as much as we wanted because it was scrapped early on but it did worked. The Arduino part is the least of your problems, getting a city wide wireless network needs an RF guy to things off the ground.We grouped nearby installations and used a backhaul to send data to the central location, 2.4ghz for the points and 5ghz for the backhaul. A local guy is handling the arduino setup, we were there for the wireless side.

Techone is right, get a couple of wireless equipment and test it at home, this is how we started. Just make sure to turn down power as it can overwhelm nearby equipment and also be mindful of your country's RF power limits. 

[Reply #9 on:]

Just a thought. If these stations are on ground level and outdoors, can you put a QR code display on their tops and have a camera pick up the info from a nearby highrise? There is no need to wirelessly connect the stations and the cameras can run dynamic routines to read each QR code display to get data from every station and upload to city hall.

[Reply #10 on:]

Basically my father in law has had an idea with a friend of his. And various municipalities has liked the idea.
Therefore he asked me if I could work on this, but I am more of an IT engineer (I'm about to graduate) and I could use this as bachelor project.

If it's too commercially sensitive to be discussed (even in abstract terms) in a technical forum then I suspect that it is not something that should be published academically. (What, you thought that everyone involved in assessing and monitoring your work would sign a non-disclosure agreement?)

[Reply #11 on:]

Basically my father in law has had an idea with a friend of his. And various municipalities has liked the idea.
Therefore he asked me if I could work on this, but I am more of an IT engineer (I'm about to graduate) and I could use this as bachelor project.

If it's too commercially sensitive to be discussed (even in abstract terms) in a technical forum then I suspect that it is not something that should be published academically. (What, you thought that everyone involved in assessing and monitoring your work would sign a non-disclosure agreement?)

No I did not think that.
I only mentioned the project because I had to give a very rough idea of what kind of device I am looking for in order to see if someone knew where I could find such a device that could fit my need.

No reason to get "heated up", no one forces no one to answer. I just asked a question, and as far as I remember this is legit, than people are welcome not to answer.

ON the other hand I appreciate everyone who has taken their time to contribute with their input. I really appreciate it.

[Reply #12 on:]

Just a thought. If these stations are on ground level and outdoors, can you put a QR code display on their tops and have a camera pick up the info from a nearby highrise? There is no need to wirelessly connect the stations and the cameras can run dynamic routines to read each QR code display to get data from every station and upload to city hall.

Hey thanks for the thought. However if i understood your idea correctly, I'd not work in my case, as this measurements performed by the sensors internally to this Objects we are developing, will have to be sent right away to the computer located in the municipality as based on some of these values some employee would have to go to that specific location.