2013-12-02

Co.Labs

Cheap Drones Are Transforming The Ancient Art Of Mapmaking

Inexpensive drones could be used to create fast, accurate maps of areas satellite imagery can’t see.



Compared to their paper-based ancestors, today's maps are pretty incredible. Yet as mind-blowing as Google Maps or OpenStreetMaps would be to a 19th-century time traveler, these maps could be even better--if they were made by drones.

In a recent demonstration, data analyst Bobby Sudekum and a few of his colleagues at MapBox sent a SenseFly eBee into the air above Lost Creek Winery in Virginia. Within less than hour, the lightweight drone collected 100 acres' worth of photographic imagery, which was then pulled into MapBox's suite of tools for editing and publishing interactive maps.

The stunt was smart fodder for tech blog coverage, but it also made a powerful point: In the future, inexpensive drones could be used to quickly and accurately create maps of areas for which there might not be very good satellite imagery--or more commonly, areas for which the data is old.

"It's going to be huge," Sudekum says of drone-powered mapmaking. "Right now, the state of drones is like 1970s in the computer world. The sky's the limit."

While the technology will certainly continue to get cheaper and more advanced, drones already really good at at least one thing: producing photorealistic maps very quickly. As handy as the maps produced by Google or Bing are, they're old--check out your old neighborhood and notice what’s changed since those satellite images were made.

Outdated maps are fine if you need directions to a store in the next town over. But for anybody interested in mapping things in a more timely manner--be it for journalism, urban planning, farming, disaster prevention, emergency response, or anything else--today's maps are too stale.

"I think drones are going to be playing a huge part in getting precise, accurate imagery. There's so much we can approve upon. There's so many awesome use cases, like agriculture and disaster relief," Sudekum says.

Last week's Virginia flyover was just the latest real-world example showing how drones can be used for making maps. In mid-October, a swarm of camera-wielding drones autonomously built a 3-D and highly detailed map of the Matterhorn. Meanwhile, ecologists interested in producing timely, three-dimensional maps of vegetation can turn to , a low-cost 3-D mapping toolkit supported by the National Science Foundation.

For an idea of how this new technology could benefit real people, look no further than Haiti. In April of this year, a team from an organization called Drone Adventures arrived in the country with three eBee drones and a stack of computers running Pix4uav, a desktop app for compiling aerial imagery and stitching it together to create 2-D and 3-D maps.

Over the course of a week, they mapped 28 square miles of land, including densely populating urban shantytowns and dangerous terrain outside the city. Like the Matterhorn and Virginia mappings, this one was more for demonstration than for practical usage. But one can imagine how handy up-to-the-hour maps would have been when an earthquake hit Haiti in 2010. There's something to be said for having recent imagery when buildings crumble and roads are suddenly in ruins.

For MapBox, drone mapping is still an experimental practice with no immediate impact on the company's product roadmap--at least, not yet. They're just exploring the possibilities for now, but are already beginning to think about where open source drone mapping might fit into their future.

"We've been thinking about extending the idea behind OpenStreetMap to a crowdsourced aerial imagery map," Sudekum says. "More and more hobbyists are adding cameras to planes and mapping their surroundings. What if they had a community network map where they could add their imagery? A worldwide crowdsourced imagery map."

[Image: Flickr user The Library of Congress]






Add New Comment

1 Comments

  • Ernie Schell

    I've always been a fan of maps, and of course in daily life need them to get around. But in the pre-Web era (going back to the 1970s), it was always a challenge to decide just when a map I already had was out of date and needed replacing. I won't rehearse the logic of it (rather convoluted), but I concluded that every three years was the interval for places I needed details about. For major highways, five years was OK, but if the trip was critical (time-sensitive, no tolerance for the unexpected), three years was the maximum there, too. Based on this article, three months may be more like the current interval for updating digitized maps. But I can easily imagine in the not-too-distant future that extremely small micro drones that are recharged every night might blanket most of the well-populated areas (except cities, because of other hazards) and update digital maps daily (not that they would need to be, of course, except for alerting drivers to temporary detours, roadwork, accidents, or heavy traffic. For the latter GPS already does a reasonable job).