Estimation: Size the Total Data Storage Behind Google Street View Imagery

Estimate how much storage Google needs to hold the Street View imagery in Google Maps. Talk through your assumptions as you build the number. ```hint Follow the car Imagery is captured at points along roads. Work out how much data one capture point produces, how many such points a road network needs, and what else multiplies that total beyond a single snapshot of each place. ``` ### Clarifying Questions - Should I estimate worldwide coverage, or size one country such as the US first and then scale up? - Am I sizing only the published panoramas that users browse, or also the raw data the capture vehicles record for processing? - Should older captures that stay available as historical imagery, and extra copies kept for durability, be included? - Is a storage volume enough, or would you also like a monthly cost? ### What a Strong Answer Covers - A clear equation that multiplies road coverage, capture density, data per capture, retained versions and redundancy - Round-number assumptions for each factor, each labeled and justified, with units kept straight (road miles versus square miles, megabytes versus megabits) - Correct arithmetic, with the result stated in sensible units - A sanity check against an independent reference point, and a note on which assumption moves the answer most - How the estimate extends to raw capture data, other countries or imagery away from roads ### Follow-up Questions - Which assumption would you check first, and how would you check it? - How would the answer change if every raw camera frame were kept instead of only the stitched panoramas? - How could you cut the storage bill without making Street View worse for users?