# How accelerometers sense constant velocity movements

There is something I don't understand about accelerometers. I know it's possible to make an Inertial Measurement Unit (IMU) by using "three" accelerometers. So with that, I could calculate the $x$, $y$, $z$ coordinates of something in 3-D space.

The point that I don't understand is, how could it sense constant-speed movements? When the velocity is constant, acceleration is zero. So how does it work?

-

It can't directly(Einstein was right)

What you can do is integrate all the accelerations you have measured upto that point in time to have a current knowledge of the current velocity.

This isn't very accurate because any small inaccuracies in acceleration will lead to increasingly inaccurate velocity, and so position, with time - this is known as drift.

-