Before Google and Apple provided free navigation apps we had personal navigation devices and builtin car navigation systems with comprehensive maps. The revenue from hardware sales and map updates supported two multi billion companies, Navteq and Teleatlas, that compiled map data for the most developed countries. They did drive around everywhere with two people in each data collection car, a driver and a data collection person making voice and text notes. The notes were later computer and human processed to the final digital map database.
Google basically did the same but replaced the data collection person with video recorders and automated recognition of road signs and house numbers. Some hard to recognize house numbers and street names are automatically recognized with [CAPTCHAs](https://en.wikipedia.org/wiki/CAPTCHA). Google presents the same image to many people. If most of them recognize the same number or letter that’s considered a valid recognition result. Besides that Google had [Google Map Maker](https://en.wikipedia.org/wiki/Google_Map_Maker) program open to absolutely everybody. Once they started to dominate in some countries they were able to buy manually collected data from companies like Navteq and Teleatlas because the business of map data collection started to collapse. Apple bought initial data from Teleatlas.
Google and Apple still have to drive around to continue improving data but they can be smart about where and when they drive. They can ask users if data is wrong or detect errors automatically for example if users don’t follow navigation guidance.
Latest Answers