Phone sensors

[webber]

Is there a way to determine the actual acceleration of device (increase of speed in time) using sensors?

A way could be reading variations of speed via GPS, but if possible I'd prefere using sensors.

Thank You

 

[Erel]

Yes: http://www.b4x.com/forum/basic4android-getting-started-tutorials/6647-orientation-accelerometer.html
And: Basic4android - Phone

 

[webber]

Sensors

I'm sorry Erel, but I think it isn't so.
I reckon taht the accelerometer values returned by the function are absolute values representing orientation of device on the three axis X,Y,Z.

In fact, using the software sample if you keep the device steady in one arbitrary position, values of X,Y and Z represent the position.

The value I'd need is acceleration, i.e. the rate of increase (or decrease) of velocity.

I thought I could read the GPS speed at a given interval and calculate the rate of increase/decrease but wanted to know if accelerometer could have been useful to the scope as I think GPS increases power consumption.
Bye, and thanks again.

 

[agraham]

No, I think Erel is right. Orientation and accelerometer values are different. If you look at the screen shot in the tutorial X and Y are small values while Z is large so Z is measuring the earths gravitational pull on the accelerometer in what looks like m/s^2 as the nominal value is about 9.8 m/s^2. Orientation is the vector resolution of the accelerometer values assuming the device is stationary or in constant motion with respect to the earths gravitational field.

I haven't played with it but I suspect that the accelerometer is too crappy and noisy to provide anything like the accuracy that you would probably want. Integrating successive readings to get velocity change over an interval might improve the noisiness of the signal but from experience on a Windows Mobile phone the axis to axis matching and zero point settings of these cheap solid state accelerometers is abysmally inaccurate.

 

[webber]

Sensors

Thank you Erel and Agraham, I will make some attempt to see how to solve.

 

[klaus]

I confirm agraham's comments.

Trying to use the accelerometer of the device for routing, you can forget it.
- the accuracy is too low.
- expecting to integrate the acceleration to get the speed is an illusion, accumulation of the errors of the accelerometer.
- you cannot know how the device is oriented in space to determine in what direction are the x, y and z coordinates of the accelerometer.

But nevertheless, you can try it, and finaly you will confirm the statements above.

In my professional life I did lots of measurements in mechanical engineering, we had once tried to integrate accelerations, NO success. Even with precise accelerometers.

Best regards.

 

[Cableguy]

From my short experience with my device, the G-sensor are too sensible, so ANY electromag field can induce an error on the readings...Including any metal objects, power lines...whatever. I thinks the best shot for getting acceleration would be to use GPS data

 

[webber]

Sensors

Now I have to say I absolutely agree.

I have done many attempts with accelerometer to 'touch with my hands' the issue and now I think that GPS is the valid solution.

Thanks to everyone for helping.

 

[klaus]

I have done many attempts with accelerometer to 'touch with my hands' the issue and now I think that GPS is the valid solution.

This is called experience .
The best experience is the one you aquire on your own. That's why I suggested you to try also yourself.

Best regards.