Crowdsourcing Maps is Still Hard 1

Crowdsourcing Maps is Still Hard

Posted Friday, Jun 28, 2019 by Christopher Wilson

I’ve been working on crowdsourced mapping for over 20 years.  It seems like such a simple idea would be able to make the maps we need for Automated Vehicles, but it turns out it’s really hard.  A few years ago this was a very hot space for investment, but people are now realizing it will take a little longer, both for the maps and the market…   — Christopher Wilson

AI and crowdsourcing fueling mapping innovation to meet smart city and mobility needs

 

An image from Mapillary Vistas Dataset, a pixel-accurate annotated street-level imagery dataset for  autonomous mobility and transport.

Above: An image from Mapillary Vistas Dataset, a pixel-accurate annotated street-level imagery dataset for autonomous mobility and transport.

Google and Apple loom so large over the field of digital mapping that it’s understandable why it may seem they represent the beginning and the end of this market. But the demands of a wide range of services such autonomous vehicles and smart cities are giving rise to a new generation of mapping competitors who are pushing the boundaries of innovation.

The fundamental approach to mapping used by the two giants, mixing satellite imagery and fleets of cars roaming the streets, is becoming archaic and too slow to meet the fast-moving needs of businesses in areas like ecommerce, drones, and forms of mobility. These services often have very specific needs that require real-time updates and far richer data.

To address these challenges, new mapping companies are turning to artificial intelligence and crowdsourcing, among other things, to deliver far more complex geodata. This increasing diversity and competition is the catalyst behind a global mapping market that is growing more than 11% annually and is expected to be worth $8.76 billion by 2025, according to Grand View Research.

“It’s a very exciting time to be building a mapping company, because the world is getting connected,” said Alex Barth, head of auto for San Francisco-based Mapbox. “And it creates all sorts of new ways of thinking about location.”

Read more…


No Comments