A self-contained battery powered system is required that measures the number of steps taken by a person. The sensor used to detect the steps must be solid state and a 3D accelerometer is likely to be a satisfactory method. When a person is walking there is an up and down acceleration as well as a forwards and backwards acceleration. A micro controller system such as the BS2-IC module or PIC microcontroller might be suitable for processing the information from the 3D accelerometer to measure the number of steps taken. It will also need to be capable of operating when a person is walking up and down stairs, jogging and running and to be able to distinguish between these different states of movement. The system must be capable of proving information to the user on how much daily activity has been undertaken and perhaps even provide stimulus to encourage them to move more if they have been sedentary for some time. This could be in the form of beeps of various frequencies and durations. There is also potential for use in training by monitoring average step frequency and beeping if the step frequency is too low or too high, depending on the fitness requirement. This will require a suitable user interface and it would be beneficial if a USB communication link was also implemented.


Is there anyone who can please give a clue on how to go about this project as i have done the tilt test on it but do not know how or where to go from there?

 

A self-contained battery powered system is required

Is this an assignment?

 

it is a project in order words yes!

 

Since your project is a project based on a class assignment, it is more appropriate that it reside in the homework section.

One way to start a project of this complexity is to draw a block diagram of the basic circuit blocks. From the diagram you can then begin to fill in the details on a block by block basis until you end up with a full schematic.

Once you have a rough block diagram you can post it here for member comments.

hgmjr

 

If you have a droid or iphone, run a program that shows the accelerometer values. That will give you an idea of the range needed for deciding when "one pace" has been made.

It does get a bit complicated, but isn't impossible, unless it is to be the size of a wristwach.

 

How about if you purchase one of the cheap off-the-shelf pedometers and then modify it to replace the primitive bobbing weight with your own micro-based accelerometer. Or have you already determined that such a solution is strictly prohibited?

hgmjr

 

i already have a DE-ACCM3D PEDOMETER and done the test on it calculating the voltage dissipated from each pins, also i have a schematic diagram displayin what i think should be done. but honstly i am lost right after that and was wondering if anyone could advise on what next step to take.

below is an attached doc file showing the schematic for the circuit built having in mind that i am using a pic18f1320 which still needs to be programmed using 'C' to tell x, y, z what to do.

 

We are kind of allergic to MS DOC files. A graphic file format like PNG is much to be preferred.

Is that a schematic, or a flow chart? At this point, you would probably get more mileage out of a flow chart.

 

here is the .PNG file to see how far i have gone.. thank you guys for ur help so far!

 

Are you sure you need a three axis accelerometer?

One of the things we have suggested is a flow chart. That is a map of the program flow. You will need one to see how you want to write your code. You have a lot of functions to accomplish, and many of them will depend on some of the others.

 

and how do i go about the flow chart and what should be implemented within it?, because right at that point i feel at a cross road.

 

Got MS Office? - http://office.microsoft.com/en-us/w...rts-with-word-and-powerpoint-HA001055266.aspx

Another - http://spreadsheets.about.com/od/tipsandfaqs/f/flowchart.htm

Open Office - http://www.marccarson.com/site/resources/27/free-openoffice-template-flowchart

Or you can usually find a physical template to drew the shapes with at an office supply store.

A micro controller system such as the BS2-IC module or PIC microcontroller might be suitable for processing the information from the 3D accelerometer to measure the number of steps taken. It will also need to be capable of operating when a person is walking up and down stairs, jogging and running and to be able to distinguish between these different states of movemen

Collecting the data means continuously taking in conversions fro the accelerometer and storing into memory. That gives you a pile of numbers that have to be evaluated as to the type of motion (walking, running) and stuff like motion up and down stairs of slopes.

One place to start is to describe the conversion and storage process, plus whatever evaluation must be done as the exercise goes on. What depth of memory is needed to evaluate each activity and track total distance, steps, and level of exertion. How do you keep a record of past performance to determine if the user is doing more or less exertion?

right at that point i feel at a cross road

What might that be?

 

yes i do need the 3 axis accelerometer and it is serves as the major objective to the project. i have tried various ways to get the flow chart going but all to no avail.. is there any helpin hand you could lend. please?

 

Should go like this (processor thinking), no actual code:

Am I moving?
If I am not moving, take a nap to save battery.
If I am moving:
Which way is Up?
Am I moving Up or down?
Did I just change direction?
Or was I dropped on the floor?
If I changed direction, tell human he is doing good.
If acceleration is 1 in a static plane, I was dropped, scream (beep) or nap.
Am I moving up or down?
If I am moving up, keep waiting for me to go down.
If I am moving down, am I moving in generally the same plane and speed as a human stride?
If not, yell at human for shaking me in order to cheat.
If so, keep counting and telling the human he is a good person.
Repeat until battery dies.

That should get you started on a flow chart. Circles are start and end, decisions diamonds, rectangles are tasks (or something like that), with lines and arrows and stuff to make it look all professional.

--ETA: You'll need to "flesh it out" a bit, detecting rest periods, maybe measuring pulse or something if you go advanced, keep a "score" system of points in EEPROM, and the best way to do it is to THINK LIKE THE PROCESSOR! That is the best way to make an interactive device (pedometer, robot, car, etc) more functional and less buggy. Think of all the other contingencies, plot them out, then code it.