A PING cannot measure phase shift. All it can do is produce a pulse of ultrasonic sound and measure the time to the first received echo to give an estimate of the distance to the nearest object in the field of view.
If you could measure phase shift, that would give you a relatively fine-grained distance measurement, not a measure of speed. Frequency shift would give you a measure of the Doppler effect which would give you an estimate of speed. Neither would give you direction information. To use phase shift to infer direction, you need multiple readings across a plane. The PING is a single transmitter / receiver. You'd need a receiver array.
Re-reading your response, I'm thinking you may have thought I was thinking of "bearing" when I said direction. All I want to do is measure if an object is moving towards or away from me. I'd guess you could compare distance changes as going +/- ? But thought there might be a more elegant method?
The only thing you can do, as you suggested, is to take successive distance measurements at intervals and see if (and by how much) the measurements are changing. If the object is moving slowly enough to get a series of measurements, you would be able to estimate the speed of the object towards or away from the PING.
Comments
If you could measure phase shift, that would give you a relatively fine-grained distance measurement, not a measure of speed. Frequency shift would give you a measure of the Doppler effect which would give you an estimate of speed. Neither would give you direction information. To use phase shift to infer direction, you need multiple readings across a plane. The PING is a single transmitter / receiver. You'd need a receiver array.