Descartes Labs launches its geospatial analysis platform

Descartes Labs, a New Mexico-based geospatial analytics startup, today announced that its platform is now out of beta. The well-funded company already allowed businesses to analyze satellite imagery it pulls in from NASA and ESA and build predictive models based on this data, but starting today, it is adding both weather data to its library, as […]

Descartes Labs, a New Mexico-based geospatial analytics startup, today announced that its platform is now out of beta. The well-funded company already allowed businesses to analyze satellite imagery it pulls in from NASA and ESA and build predictive models based on this data, but starting today, it is adding both weather data to its library, as well as commercial high-resolution imagery thanks to a new partnership with Airbus’ OneAtlas project.

As Descartes Labs co-founder Mark Johnson, who you may remember from Zite, told me, the team now regularly pulls in 100 terabytes of new data every day. The company’s clients then use this data to predict the growth of crops, for example. And while Descartes Labs can’t disclose most of its clients, Johnson told me that Cargill and teams at Los Alamos National Labs are among its users.

While anybody could theoretically access the same data and spin up thousands of compute nodes to analyze it and build models, the value of a service like this is very much about abstracting all of that work away and letting developers and analysts focus on what they do best.

“If you look at the early beta customers of the system, typically it’s a company that has some kind of geospatial expertise,” Johnson told me. “Oftentimes, they’re collecting data of their own and their primary challenge is that the folks on their team who ought to be spending all their time doing science on the datasets — the majority of their time, sometimes 80 plus percent of their time — they are collecting the data, cleaning the data, getting the data analysis ready. So only a small percentage of their work time is spent on analysis.”

So far, Descartes Labs’ infrastructure, which mostly runs on the Google Cloud Platform, has processed over 11 petabytes of compressed data. Thanks to the partnership with Airbus, it’s now also getting very high-resolution data for its users. While some of the free data from the Landsat satellites, for example, have a resolution of 30m per pixel, the Airbus data comes in at 1.5m per pixel across the entire world and 50cm per pixel over 2,600 cities. Add NOAA’s global weather data to this, and it’s easy to imagine what kind of models developers could build based on all of this information.

Many users, Johnson tells me, also bring their own data to the service to build better models or see

While Descartes Labs’ early focus was on developers, it’s worth noting that the team has now also built a viewer that allows any user (who pays for the service) to work with the base map and add layers of additional information on top.

Johnson tells me that the team plans to add more datasets over time, though the focus of the service will always remain on spatial data.

UberAIR to take flight with help from UT Austin and U.S. Army Research Labs

After three months of discussions, Uber Elevate has selected The University of Texas at Austin as its partner alongside the U.S. Army Research Laboratory to develop new rotor technology for vehicles that the company will use in its uberAIR flying taxi network. The news is the latest step in Uber’s plans to get demonstration flights […]

After three months of discussions, Uber Elevate has selected The University of Texas at Austin as its partner alongside the U.S. Army Research Laboratory to develop new rotor technology for vehicles that the company will use in its uberAIR flying taxi network.

The news is the latest step in Uber’s plans to get demonstration flights off the ground in the megalopolises of Dallas Ft. Worth; Los Angeles, and Dubai. The ultimate goal is to have uberAIR services commercially available in those cities by 2023.

To achieve that, Uber has set up some rigorous specifications for its vehicle and the traffic management system used to operate uberAIR, developed in conjunction with several aircraft manufacturers and the National Aeronautics and Space Administration.

Specifically for the vehicle, Uber is requiring a fully electric vertical take-off and landing vehicle that has a cruising speed of 150 to 200 miles per hour; a cruising altitude of 1,000 to 2,000 feet; and a range of up to 60 miles for a single charge.

The company isn’t the only one racing to own the sky taxi space for urban transport. Chinese drone manufacturer Ehang; Aston Martin; Rolls Royce; Audi and Airbus and other, smaller, startup vendors are all trying to make flying vehicles. Ehang has been touting manned test flights of its drone already.

Uber, on the other hand is trying to build out the service in much the same way it did with car hailing so many years ago.

The company actually unveiled its thoughts on air travel and design a few months ago at its Elevate conference.

At UT, a research team led by Professor Jayant Sirohi, one of the country’s experts on unmanned drone technology, VTOL aircraft, and fixed- and rotary-wing elasticity will examine how the efficacy of a new flying technology, which uses two rotor systems stacked on top of one another and rotating in the same direction.

Called co-rotating rotors, the new technology will be tested for its efficiency and noise signature, according to a statement from the university. Preliminary tests have shown the potential for these rotors to work better than other approaches while also improving versatility for an aircraft.

“There’s a lot of things to be done,” said Sirohi. “We are not doing vehicles. we’re doing a specific rotor system on one of the engineering common reference models that Uber has released.”

The reference model is a benchmark for what the aircraft should do in field tests and eventually operations, Sirohi said. “We are pursuing these technologies to see what the gaps are in where we are today and where we need to be,” Sirohi said.

Boeing’s new R&D center focuses on autonomous flight

Flying cars are BS. But there is actually a chance that we’re on the cusp of a revolution in general aviation as startups and major players like Airbus are looking to modern technology to allow more people to take flight without having to first learn how to steer a Cessna 152 down a short runway […]

Flying cars are BS. But there is actually a chance that we’re on the cusp of a revolution in general aviation as startups and major players like Airbus are looking to modern technology to allow more people to take flight without having to first learn how to steer a Cessna 152 down a short runway (though that’s a good skill to have, too). Boeing, which is not currently a player in general aviation, clearly doesn’t want to be left behind. The company today announced that it is opening a new R&D office in Boston that will focus on designing, building and flying autonomous aircraft.

Perkins + Will (PRNewsfoto/Boeing)

The new office will be staffed by Boeing engineers and employees of its Aurora Flight Sciences subsidiary, which it acquired last year. Unsurprisingly, the company also plans to work with the boffins at MIT, which is the landlord of the company’s new offices in Kendall Square.

“Boeing is leading the development of new autonomous vehicles and future transportation systems that will bring flight closer to home,” said Greg Hyslop, Boeing’s chief technology officer, in today’s announcement. “By investing in this new research facility, we are creating a hub where our engineers can collaborate with other Boeing engineers and research partners around the world and leverage the Cambridge innovation ecosystem.”

There’s obviously plenty left to be figured out with regards to autonomous flights — or even just giving people access to semi-autonomous aircraft that would fall under the FAA’s ultralight designation so that the pilot wouldn’t need a pilot’s license.

Right now, the rules are pretty clear about what’s possible and what isn’t — and most ideas around “urban air transportation” (at least in the U.S.) aren’t feasible under today’s rules. But those rules were written for an aviation system that handles fewer than 250,000 small general aviation aircraft, most of which feature very little in terms of automation. The FAA has shown some ability to act relatively fast when new technology comes along (see: drone regulations), so maybe that’ll be the case here, too. By 2020, the average Cessna you see puttering about in the sky above you will be close to 50 years old, but who knows, maybe we’ll see sleek electrified personal Boeing drones zooming 500 feet above our heads instead.