To madhavank007, you need pressure differences. If you can introduce a known flow "resistance" such as a valve, an orifice, a venturi nozzle, or a pitot tube, and measure the pressure drop across this resistance by suitably located taps, you may estimate the flow rate from the velocities therein, based on Bernouilli's theorem. Nothing new, however. These are methods used to measure flow rates in pipes. Rotameters can also be used.
There are formulas to estimate flow rates by letting the fluid exit the pipe to the atmosphere or flowing over a notched weir. By measuring distances (vertical and horizontal, as the case may be) from the open nozzle or weir, one can get a reasonable flow rate appraisal.
Once you can measure flows, you may construct a table based on pump discharge pressures, keeping the piping geometry and head at the receiving point unchanged, to relate these pressure readings to measured liquid flowrates.
To quark, it has been shown that the Babylonians used tables of squares to solve multiplications based on the formula:
a.b=[(a+b)
2-(a-b)
2]/4
and that the old egyptians and chinese used the 3,4,5 "rule" to construct right angle gardens and buildings.
Amazing feats by themselves, thousands of years before Newton's famous binomial and Pythagoras' theorem proof.