Two Different Ways to Make Maps for Self-Driving Cars

Another piece on the various attempts to create detailed, high-definition maps for self-driving cars, this time from Bloomberg’s Mark Bergen, who views it through the prism of Google’s efforts in that space, and whether its competitors will be able to stop Google from dominating the high-definition mapping space the way it has come to dominate consumer maps.

There are, Bergen reports, two ways to make high-definition maps for self-driving cars:

The companies working on maps for autonomous vehicles are taking two different approaches. One aims to create complete high-definition maps that will let the driverless cars of the future navigate all on their own; another creates maps piece-by-piece, using sensors in today’s vehicles that will allow cars to gradually automate more and more parts of driving.

Alphabet is trying both approaches. A team inside Google is working on a 3-D mapping project that it may license to automakers, according to four people familiar with its plans, which have not previously been reported. This mapping service is different than the high-definition maps that Waymo, another Alphabet unit, is creating for its autonomous vehicles.

China Restricts Foreign Firms from Mapping Roads for Self-Driving Cars

Self-driving cars require insanely detailed maps in order to function. But, as The Drive’s Stephen Edelstein writes, “The Chinese government is blocking foreign companies from mapping its roads in great detail, according to a Financial Times report. The restrictions, which reportedly do not apply to Chinese firms, are being instituted in the name of national security. China is concerned about spying.” Mapping, geotagging, geographic surveys—all of these have been subject to Chinese government restrictions for many years (recall the trouble Google Maps has had operating in China), so this is more of an additional data point than an actual surprise. [Boing Boing/PC Mag]

Previously: The Business of Making Maps for Self-Driving Cars.

The Future of Mapping: Semantic Maps

The future of mapping, according to John Hanke and Brian McClendon, who basically created Google Earth, is something called a semantic map. Now, semantic mapping has very different meanings in literacy, statistics and data science; in this case a semantic map refers to “a map that continues to learn about the physical world and refine its predictions of what objects are and how they will act through huge amounts of data. Without semantic maps, self-driving cars won’t ever be able to intelligently move through the world–at least not without crashing into something—a development that will arrive when self-driving cars do.” Also has augmented reality implications. The Co.Design article explains. [Dave Smith]

The Business of Making Maps for Self-Driving Cars

CNN on the big business involved in creating detailed maps—called HD maps—for self-driving cars. “If you believe self-driving cars will eventually operate everywhere, then every city and street will need to be mapped out in granular detail.” How granular? During one test, a single-pixel error on one map caused cars to avoid a patch of road as though it was raised 10 inches. [Osher]

Previously: Human-Annotated Maps for Self-Driving Cars.

Human-Annotated Maps for Self-Driving Cars

Self-driving cars need extremely detailed and comprehensive maps in order to work—far more detailed than what’s usually available. Paradoxically, Vox’s Timothy B. Lee reports, that’s going to require significant human labour, in the form of human analysts annotating the map. “As Google and its competitors expand their self-driving vehicle programs nationwide, they’re going to have to hire thousands of human analysts to produce the detailed maps that enable cars to drive safely.” [MAPS-L]