Automakers and tech firms want high-definition maps to help robo-cars drive, but the best way to build them remains unclear. So says a new report from Bloomberg.

Two approaches: “One aims to create complete high-definition maps that will let the driverless cars of the future navigate all on their own,” explains Bloomberg. “Another creates maps piece-by-piece, using sensors in today’s vehicles.”

But: Those approaches contain scope to do things differently. “Every self-driving map looks different because each one depends on the sensor system of the vehicle that creates it,” says Bloomberg.

Many players: Waymo is thought to be leading the way, but other firms—including Mobileye, Tesla, and TomTom—are in the scrum. Plus, says Bloomberg, a “slew” of startups are “taking different stabs at the problem, each gobbling up venture capital.” All use subtly different approaches.

Winner takes all? Andreessen-Horowitz partner Benedict Evans says these maps have a network effect: a bigger fleet of cars using one type of maps will get better, more regularly updated maps faster than a smaller one would. So there’s plenty to play for.