The route itself not the most complex calculation, but true complexity resides in that how do you make the result is OK with both driver, rider and uber. All k

The route itself not the most complex calculation, but true complexity resides in that how do you make the result is OK with both driver, rider and uber. All kinds of things happen as you don't know the preferences of everyone in the system (does the rider want to get to the place as fast as possible, or as cheaply as possible?) and because choosing the most optimal route has random variance (stop lights, traffic etc.) Also, there are multiple different pricing models at work constantly, at the same time, in different areas. Some country might use a "taxi-meter-simulator" type methodology where the actual route of the driver is mapped via constant GPS and other positioning technology. This introduces the randomness of a certain % of GPS errors, which need to be corrected manually. Other option might be a fixed price that is estimated upfront and agreed upon by both rider and driver. Or sometimes the the fixed price is only agreed upon by the rider, and the driver gets paid via the "taximeter-style" method. And after all this, if you want to take into account things like how to translate a pin position or restaurant name into the best possible GPS position? How to optimize what driver is the best to pickup the passenger? The lower rated driver that is close or the 5 star that is far? Can we change optimal drivers on the fly? And it goes on and on, deeper and deeper.