Revamped Apple Maps Better than Google Maps but Accuracy and Coverage Needs to Improve

Apple Maps Flyover Featured

Apple announced a major revamp to Apple Maps earlier this year. Back then, Tim Cook claimed that “there’s no one doing this level of work”. So, is Apple’s refreshed map data any better than Google Maps?

As per a detailed comparison from Justin O’Beirne, Apple seems to have made a lot of lost ground in terms of maps data when compared to Google Maps. There’s a “staggering amount of vegetation detail” now, with Apple also focusing on many of the smaller cities and counties. When compared to older map data, the difference is particularly striking, with nearly 25 percent of the county seats having no vegetation data previously.

The level of vegetation captured by Apple Maps is simply astounding. It is far more than TomTom, Waze, and even Google Maps.

The impressive details are limited not just to vegetation. Apple Maps now has more data about harbors, racetracks, parking lots, football fields, and more.

Apple is also upgrading the building data footprint with more details that were simply missing from the older maps data.

But there’s a catch — there’s a wide variation in building data and some glaring error to go along with it. Apple has been extracting building and just about any identifiable shape data from it satellite imagery and then adding it to maps.

This approach is similar to that of Google. However, the wide difference in building data points to something else as well. There is a difference in building data like new Apple Maps data erroneously showing Four Embarcadero Center and Five Embarcadero Center as almost the same height, though the latter is almost twice as tall.

Justin O’Beirne says that while Apple is using lidar data captured from its vehicles running around the United States, it is using its Maps development office in Hyderabad, India for manual creation of building data. This data is further corroborated by the high headcount of the office at around 5,000 people.

He blames the manual creation of building data by different modelers for their height inconsistencies and poor modeling. Plus, given that the new Maps data has very limited building data, the theory of Apple using a considerable amount of manual labor to improve Apple Maps makes sense.

This also explains that despite Apple working on the new maps data for over four years, its coverage is so poor. The new map data covers just about 3.1 percent of the entire U.S. and around 4.9 percent of its total population.

Make sure to read the full in-depth comparison by hitting the source link below since it provides more insight into how Apple has managed to improve its maps data and exceed that of Google Maps in some scenarios.

[Via Justinobierne]