Justin O’Beirne recently wrote a blog post surveying the landscape of mapping apps and sites (with particular focus on Google Maps and Apple Maps). This is one of a series of pieces O’Beirne has written comparing Google and Apple Maps. In short, O’Beirne concludes that Google has a substantial competitive advantage in mapping that boils down to having better data.
The article is centered around the example of Google’s Areas of Interest feature. Simply put, Google adds subtle shading around areas that are commercial districts, i.e., clusters of restaurants and shops, while Apple does not. This feature is quite useful as it reflects the way humans think of cities in terms of their commercial districts.
Court and Smith Streets are commercial districts in the Cobble Hill neighborhood of Brooklyn
Google built this feature from two distinct datasets: satellite imagery and street view imagery. Apple has - or had - access to similar datasets, but Google outpaces Apple in both categories to great effect.
Google uses machine learning and computer vision to digest huge volumes of satellite imagery into 3D outlines of buildings with great detail, and they do it automatically, resulting in tremendous coverage within the United States with very high accuracy. Meanwhile, Apple appears to buy its building data from outside vendors.
Google also uses machine learning to process its street view imagery to understand which businesses are where. This provides very precise positioning of businesses that allow Google to define the boundaries of Areas of Interest with great accuracy.
Applying machine learning to these two datasets gives Google a superior product to begin with: more coverage of (and more accurate) building models, and more accurate building locations. But combining these datasets yields a further advantage. Google knows exactly where buildings are and what they contain, so they can automatically define Areas of Interest. Apple simply lacks the data to do so.
In sum, Google’s applications of machine learning and access to superior data make possible a useful feature that Apple just cannot compete with in a meaningful way.
More from the Blog
Jan 22 2018
by — While we commonly associate recommendation systems with e-commerce, their application extends to any decision-making problem which requires pairing two types of things together. To understand why recommenders don’t always work as well as we’d like them to, we set out to build some basic recommendation systems using publicly available data. Data The first ingredient for building a recommendati...
Jan 26 2018
Cloud computing has transformed enterprise IT. It makes it possible to rapidly scale up and down a complex and global infrastructure, without the overhead of a datacenter. But living in the cloud can be expensive, and you still need to maintain computers and their operating systems, even if they are virtual. That’s why we’ve been watching with interest the rise of “serverless” computing. Serve...
Apr 3 2019
by — Many interesting learning problems exist in places where labeled data is limited. As such, much thought has been spent on how best to learn from limited labeled data. One obvious answer is simply to collect more data. That is valid, but for some applications, data is difficult or expensive to collect. If we will collect more data, we ought at least be smart about the data we collect. This motiv...