Indoor Positioning system using Arduino

I want to create a Indoor positioning system using arduino and active rfid chip with range… say 50 meters.
i want to use a 2.4ghz rfid ic for transmitting and receiving the data with single tag and three readers. by using triangulation algorithm and position measurement i think i can find the location of object.

But i learned from google that nrf24lo1 doesn’t have an RSSI indicator. so there is no distance calculation using this IC.

so i gone through cc2500 which i think meets my requirements.

can anyone give me any kickstart references to interfacing cc2500 with arduino uno/mega. and building a reader using arduino and cc2500.

and please post the references if any.
i try to google it but i didnt have any luck.

This has been discussed many times before on the forum, and its not technically possible using the sort of technology you mentioned.

It requires specialist hardware that can measure the time differences of signals arriving from different devices, to incredible accuracy.

Aside from that. RFID readers don’t have a range of 50m, its more like 2 or 3 m absolute max.
But perhaps you meant the cc2500, but even then, 50m is a bit of a stretch. I’ve used other 2.4Ghz devices and have found they have a very limited range. It could be 50m in the middle of a football field, but in a normal domestic environment, they are limited to communicating with devices in the same room, as the signals don’t go though brick or other walls very effectively.

hi, rogerClark,

Thanks for the reply.

Is it possible to find the distance from one rfid device to another using RSSI? and about the ranges please see the link http://skyrfid.com/RFID_Tag_Read_Ranges.php

Signal strength (RSSI) is not a good way to determine distance from the transmitter, there are too may factors apart from distance that effect RSSI including reflections from objects, and differences in orientation of the transmitter antennas with respect to the receiver etc.

The only accurate method to determine position is what GPS does, which is measures the time taken for the signal to get between the satellite and the GPS receiver. The GPS satellites have atomic clocks on them and the GPS chips have dedicated hardware to measure time delays etc.

Its not something you are going to be able to build yourself.

Is there anyway to make Indoor Positioning system.
RSSI Isn’t the accurate way to tell a position. But we can tell the position of tag within Indoor open space like you said in football field.

No.

Using RSSI will not give you very good results even in an open space. The best it would give you is possibly which "tag" was closest.

You also have another issue, if "tag" means passive RFID tag or Near Field device. As the range of these devices is only a few metres at most.

It is possible to build a system that would give you a position of a "tag" but it would either require that the tags are "active" and have a battery or power supply so they can transmit data when required.

You would also need to use very accurate time keeping systems, requiring dedicated digital hardware, and possibly the use of atomic clocks.

If you had enough receivers and carried a passive tag or enough passive tags and carried a receiver then you could pin position down very closely. Operative word is 'enough'.

i want to track a active tag with multiple rfid receivers.
i think cc2500 will do the both tag and receiver functions.
with a single tag , three readers and triangulation using the rssi are enough?

I very much doubt it would work, but you are welcome to try ;-)

Rogerclark,

I just Read that the GPS Receivers works on quartz clock But the satellites use atomic clock for accurate positioning.

So Why Don't I try To Implement same using Quartz clock using RFID? although I don't need Pin Point Accuracy?

It doesn't do any harm to try things, as you will get to understand why this will never work using RSSI.

anilkunchalaece: Rogerclark,

I just Read that the GPS Receivers works on quartz clock But the satellites use atomic clock for accurate positioning.

So Why Don't I try To Implement same using Quartz clock using RFID? although I don't need Pin Point Accuracy?

Maybe because satellites are huge distances away from you and each other and light travels about 300m per microsecond?

You could try using sound or sound and light. Flash an IR led 'trigger' then emit a sound pulse, requires line of sight to at least 3 pickups. For practical purposes the light is instant while sound takes time. Be sure to include room temperature in speed of sound calculations and some way to filter out echoes -- don't 'locate' faster than the echoes will arrive.

You could put a tag or reader every 50cm in a grid under the floor and the other part on your person. Only close by reader will "see" the tag. The pattern will do the locating.

You might be able to do the same with magnets and security Hall sensors.

i try to google it but i didnt have any luck.

That is nature's way of telling you that what you are planning probably won't work.

As there have been several requests over the last couple of weeks where people wanted to build locator systems that were physically impossible to create (i.e break laws of physics to do with electric fields, propagation time measurement etc etc etc)

Just thought I'd post a link to an article on a discredited Kick Starter project which claimed to be producing a locator BlueTooth tag

http://mobile.slashdot.org/story/14/06/23/2357200/500k-energy-harvesting-kickstarter-scam-unfolding-right-now

If someone can dig out the link to the other posting where the OP was asking how to build a system virtually identical system to the Kickstarter one, it may be helpful to cross reference these ridiculous requests.

Edit. Here is a link to the actual kickstarter page. Note how this has been suspended by kickstarter themselves. https://www.kickstarter.com/projects/yuansong84/ifind-the-worlds-first-battery-free-item-locating

Have you seen the Solar Roads kickstarter? They've pulled down over 2 million including $750,000 in tax money for a scheme that can't work.

No. Not seen the Solar Roadways one.

I'm amazed that these sorts of crowd funded projects are not illegal.

Either these people are deluded optimists that believe that fundamental laws of physics etc can be broken if required, or they are snake oil sellers from the start.

What makes it more problematic, is that people see this stuff as genuine, and the post project guidance requests on this forum to mimic something they've seen raise $$$ on these crowd funding sites. And then can't believe the advice of the respondents, all of whom tell them that what they are trying to do is impossible.

I.e what's next, perhaps Arduino perpetual motion, ;-)

The atomic clocks are used to synchronise the signals from the satellites with sufficient accuracy, given that there is no other way for this to happen as they are not connected with each other. This gives an accuracy of a few metres.

You could construct an indoor (or outdoor) positioning system (and no doubt it has been done) using an RF field generated by three or more transmitters given that they were synchronised either by being directly connected or by transponding the signal from one "master". Quartz accuracy is no longer the concern.

It becomes instantly clear that such a system is not based on Arduinox or commonly available RF modules, but requires design expertise in UHF radio systems and Gigahertz ASICs.

rogerClark: No. Not seen the Solar Roadways one.

I'm amazed that these sorts of crowd funded projects are not illegal.

Either these people are deluded optimists that believe that fundamental laws of physics etc can be broken if required, or they are snake oil sellers from the start.

What makes it more problematic, is that people see this stuff as genuine, and the post project guidance requests on this forum to mimic something they've seen raise $$$ on these crowd funding sites. And then can't believe the advice of the respondents, all of whom tell them that what they are trying to do is impossible.

I.e what's next, perhaps Arduino perpetual motion, ;-)

There would be less snake oil if there weren't crowds willing to hand over cash for gold-painted-manure.

You can check out the Youtube videos debunking Solar Roads. They're already rich as long as they can keep from being charged with fraud, which probably won't happen.

Due the high speed of radio transmissions and therefore the low margin for error in timing I think the OP will run into problems with achieving acceptable accuracy.

A slower propagating wave such as sound might be a better option and also the primary elements are already available. Without a description of the end use of the system it is hard to tell whether the compromises involved would be a deal breaker. I make the assumption that the OP wants to use the location system for robots in a room. Therefore I ask if the 50 metres is overkill. The system I propose may would be limited to a room/area somewhere beyond 4 metres and maybe beyond this (see explanation later).

An ultrasonic sensor (e.g. SR-06 or better spec unit) would be placed in each corner of the room. Each of these units would be connected to their own, dedicated arduino (or simple Atmega85/168/328 chip) with an infrared sensor also connected, through appropriate supporting circuitry, to an interupt pin of the Atmega. The robot somewhere in the room would have two or more SR-06 units to accept the pings that will come to it. When an infared pulse is periodically sent from the robot to the coner modules they would fire off a ping from their SR-06 and would ignore the reponse. To avoid the collision of pings there would be a set delay of say 100ms between corner modules. e.g. corner unit #1 one pings instantaneously on receipt of pulse, corner unit #2 100 ms after receipt of pulse, corner unit #3 200 ms after receipt of pulse, corner unit #4 300 ms after receipt of pulse. these delays would be accounted for in the final calculation in the robots micro.

The robot would receive these pings, and triangulate it position from the timing of these pings. There would be a propagation delay involved in sending, receiving and and processing the infrared pulse before a ping was fired which would need to be calibrated. Its not as accurate as pulsing an arduino pin directly connected to the trigger pin of the SR-06. Likewise there may be a very minor difference (we are talking about the speed of light over a few metres) in the time when each corner module received the infrared pulse depending on how far the robot is from each corner unit. Maybe assembler would be required to optimise the timing.

The SR-06 is rated to 4 metres however this is the sending the ping over this distance and the very much weaker reflection over the same distance. If the ping is from device to device without the very attenuating reflection, I think much greater distances could be achieved; 10 metres plus?

If the robot makes a flash and then a sound, the sensing units only have to receive but they should be connected to a single "brain".