Safely Travel Inside a Glacier Through the Eyes of a Drone (Video)

As reported by GizModo: Given the shifting ice can suddenly close a massive crevasse that runs hundreds of feet deep into a glacier, safely exploring them is all but impossible. Unless you’ve got access to a flying drone that isn’t sent flying out of control the second it hits an obstacle.

Many drones have protective housings around their propellers to protect the blades, but Flyability’s Gimball wraps the entire craft in a specially-designed wire cage. It doesn’t just protect the drone, it also freely rotates when it encounters an obstacle so that the drone isn’t suddenly steered off course.

The resulting video is remarkably stable given how often the Gimball drone is actually crashing into the icy walls lining the narrow crevasse. You don’t even need to be an expert drone pilot in order to pilot the drone deep into this glacier. Although, you’ll probably want to practice a bit ahead of time to ensure you can also safely fly the $25,000 drone back out.

Automakers Still Have a Lot to Learn from Tesla

As reported by The Verge: First, let's get this out of the way: arguing that big automakers need to learn anything from Tesla is bold, I admit, considering Tesla's often dire financials — and the fact that when you compare it to an automotive giant like Toyota or Volkswagen, its production output is still a rounding error away from zero. Tesla has yet to prove it can survive at scale, and that won't happen until the company is taking orders for a large number of Model 3s, producing them, and meeting demand. That's two years away at a bare minimum; likely more.

But did Tesla prod the legacy auto industry to move faster on practical, mass-market electric vehicles than it may have otherwise? Perhaps. It's not unreasonable to think so, especially considering that GM has graduated from recognizing Tesla's existence to outright trolling it lately. (Realistically, though, EVs were an inevitability even without Tesla. I'd actually argue it was that inevitability that begat Tesla, not the other way around.)

Never mind EVs, though — the Bolt will be in dealerships this year, there's a new Nissan Leaf on the way, and many other practical electrics from a variety of manufacturers are in the pipeline. That domino tile has already been tipped, whether Tesla was the one to tip it or not.

But there's another area where Tesla's influence needs desperately to be felt: upgradeability.

BMW AND GM BOTH DROPPED THE BALL LAST WEEK

Last week, BMW was singled out by the secretary of transportation over its Remote Control Parking feature officially being in compliance with federal safety standards, a big deal after the company had withheld from the US market over regulatory fears. Afterward, the company told The Verge that 7 Series vehicles already sold in the US could be retrofitted with the feature, bringing it to parity with cars sold in other markets. The next day, it reversed course, citing missing hardware on the 7s that have already been sold.

Then there's GM's Super Cruise, a semi-autonomous technology akin to Tesla's Autopilot. It's debuting on the CT6 sedan and was originally slated for 2016 availability, but it has now slipped to 2017 — and GM is already saying that CT6 examples sold beforehand won't be upgradeable, CNET reports.

Meanwhile, let's take a look at Tesla's track record. Every Model S and X on the road today runs the same software version, delivered over the air. It started installing the necessary hardware to enable Autopilot a solid year before the feature was actually enabled. Model S vehicles built before that are out of luck, yes, but you have to start somewhere. GM is taking the opposite approach: it is building obsolescence into cars that aren't even on the production line yet. (The CT6 doesn't officially hit dealerships until March.) Heck, Tesla is still supporting its long-discontinued Roadster with upgrade packages.

For the sake of argument, let's give GM some leeway here and say that the design and placement of Super Cruise's sensors and related hardware aren't yet finalized, and therefore there's nothing that can be built into the car ahead of time to prepare it for a software upgrade. (I'd say that's a little lame, considering Tesla had the hardware in production in 2014, but sure, okay.) And in fairness, many automakers are offering post-sale upgrades to CarPlay and Android Auto. But we're still nowhere near the flexibility and futureproofness that Tesla has demonstrated. Just look at Tesla Motors Club's comprehensive software changelog to get a sense of how these cars are evolving over time.

UPGRADEABILITY IS 

But a car from virtually any other automaker is a time capsule. Ford has been pretty good about upgrading Sync on production cars, but that's strictly infotainment — you won't likely get a new instrument cluster UI in your 2016 Ford Fusion, unless there's some sort of weird recall. You won't get a new regenerative braking algorithm pushed to your Chevy Volt as it sits in your garage overnight. Generally speaking, your car will not be made better over time.

This kind of atomic, immutable treatment of the automobile worked 25 years ago; it even worked okay five years ago. But today, car companies are dead set on competing with CarPlay and Android Auto, insistent on owning (or at least co-owning) the control elements of the dashboard. I say that if you want to play that game, you need to play by the same rules that Apple and Google brought to the table — rules invented with the smartphone, and ported to the car — which say that these devices inherently improve over time. They have to, by their nature. At a bare minimum, connected cars, like smartphones, need to be able to respond to cybersecurity threats. And software moves quickly enough that there's no reason a car shouldn't simply get better, prettier, and more useful over the months or years that you use it.

It's not that artificially limiting this type of upgradeability incentivizes car buyers to buy new cars more frequently, either, especially as most major car companies batten down the hatches for a new era of car sharing and other alternative usage models. Quite the contrary: failing to improve a car over time will alienate its user.

THEY'RE NOT THERE YET

There does seem to be some semblance of recognition across the industry that this will become table stakes in the coming years. GM's Phil Abram, for instance, told me several months ago that his company is working toward the goal of full upgradeability. "We're going to keep building on that list of items or parts of the vehicle that are capable of doing that, and working through all of the infrastructure that needs to be in place in order to do that effectively," he told me at the time. And more automakers are adding cellular connections to their cars — Ford and Toyota are recent adoptees — which make it easier to push new software.

But they're not there yet. That's really distressing in the middle of one of the most transformative periods in the history of transportation. And in the meantime, unless you're buying a $60,000-plus Tesla, I have no reason to think that the car you own today will feel current, relevant, and state-of-the art in five years.

Amazon: Our Drones 'Will be More Like Horses Than Cars'

As reported by The Verge: Amazon's delivery drones will be "more like horses than cars," according to Paul Misener, Amazon's vice president for global public policy. Misener makes the claim in an interview with Yahoo, but he doesn't mean that his company's aerial vehicles will be covered in fur and burdened with a love for sugar — instead, he likens them to horses because they'll automatically avoid obstacles that could be a danger to them.

"Try riding a horse into a tree."

"If you have a small tree in your front yard," Misener explains, "and you want to bang your car into it for some reason, you can do that. Your spouse might not be happy with you, but you can do it. But try riding a horse into the tree. It won't do it. The horse will see the tree and go around it. Same way our drones will not run into trees, because they will know not to run into it." Misener says Amazon's prototype drones have "sense-and-avoid" technology that aims to keep them out of power lines, trees, and other obstacles that could come between them and your ordered items.

That's prototypes, plural. Amazon is working on several variations of drone at the same time, Misener says, and is likely to use different drones when the time comes to actually roll the service out for customers in different locations. "Our customers in the United States live in hot, dry, dusty areas like Phoenix," he says, "but they also live in hot, wet, rainy environments like Orlando, or up in the Colorado Rockies." Their homes, too, are different. "Some live in rural farmhouses, some live in high-rise city skyscrapers, and then everything in between, in suburban and exurban environments." Those that live in larger homes with yards can expect deliveries dropped off there if they're not home when the drone calls, but Misener says Amazon is still working on how to get you your items if you live in an apartment building.

Amazon isn't worried about people shooting drones to get their items

Some have expressed concern that the sight of a drone carrying an Amazon package would entice people along its flight path to break out a rifle and try to shoot it down, but Misener dismisses the fears. "I suppose they could shoot at trucks, too," he says, glossing over the fact that trucks — unlike drones — have human drivers that might not take kindly to being shot at. But he says that Amazon thinks its Prime Air drones will be "as normal as seeing a delivery truck driving down the street someday," reducing the novelty and stopping would-be shooters from intercepting your orders.

More of an immediate concern is the legislation stopping Amazon's drones from getting off the ground. Amazon has proposed a limit that would keep piloted aircraft above 500 feet, while leaving a window from 200 to 400 feet that would allow its Prime Air drone fleet to operate, but the government body has been dragging its feet so far. "We believe that [the FAA] must begin, in earnest, planning for the rules that are more sophisticated, that go to the kinds of operations that Amazon Prime Air will encompass," Misener says. "Other countries already are doing this," he explains. "There's no reason why the United States must be first. We hope it is."

SpaceX Successfully Launches Satellite But Rocket Explodes During Drone Ship Landing Attempt

As reported by Forbes: SpaceX successfully launched the Jason-3 satellite into orbit yesterday. However, its attempted landing of the rocket on a drone ship once again failed explosively.

Apart from the landing, the launch from a SpaceX Falcon 9 rocket from the Kennedy Space Center was flawless. It lifted off at 1:42pm ET from Vandenberg Air Force Base in California.

The Jason-3 satellite is a project of NOAA, in partnership with the Centre National d’Etudes Spatiales (CNES), European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), and NASA. Jason-3 continues a series of satellite observations of the ocean which began in 1992. The purpose of the satellite is to measure changes in the ocean’s surface. This helps scientists track sea level rise due to climate change.

In addition to tracking sea level rise, the satellite will also make possible better weather forecasting, including better tracking of hurricane intensity and changes of weather patterns caused by El Nino. It will also track decadal patterns in oceans that can be used to assist the management of fisheries and other ocean-dependent businesses.

The second part of SpaceX’s goal for this launch was to once again make an attempt to land a reusable Falcon 9 first stage onto the deck of a droneship on the ocean. In December, SpaceX was successful in landing a reusable first stage on the ground following a launch. However, it’s attempts to land on a ship so far haven’t worked.  Elon Musk's video of the landing is shown here.

Yesterday was no different, although the first stage came closer than ever. According to the company, the rocket made a successful soft landing on the ship about 1.3 meters from its target. However, one of the landing legs failed, causing the rocket to fall over and explode.

NASA Picks Three Private Space Companies to Resupply the ISS

As reported by Engadget: NASA will fund the International Space Station at least through 2024, so keeping the astronauts on it fed and experimenting is a big deal. That's why the nation's top space agency announced new resupply contracts for Orbital ATK and Elon Musk's SpaceX, two major incumbents that were joined by newcomer Sierra Nevada and its resumable, mini-spaceplane. Each company will get at least six resupply missions, but it's "likely" that NASA will buy more than 18 flights — NASA will grant additional missions based on what they and the station's spacefaring crew need at the time.

Speaking of needs, these three companies were chosen in part for their different approaches to transport. Orbital ATK (which used to be known as Orbital Sciences) has its non-recoverable Cygnus capsule, which typically houses waste from the Space Station and burns up in the atmosphere upon re-entry. SpaceX's Dragon capsule can be used to both ferry cargo to the station and bring things back down, though the capsule has to be fished out of the ocean. Then there's Sierra Nevada, whose Dream Chaser shuttle can land on traditional runways — that means samples returning to earth can be offloaded and studied that much sooner. Recovery of cargo could take as little as three hours, though that all depends on whether the Dream Chaser passes its test flights later this year.

US Government Announces $4 Billion Self-Driving Car Program

As reported by Engadget: Turns out we're way closer to our self-driving car future than most of us expected. US Department of Transportation Secretary Anthony Foxx announced at the Detroit Auto Show (pdf) Thursday that the Obama administration will have a national blueprint for autonomous vehicle standards by July. What's more the administration is earmarking $4 billion of the 2017 budget to create a decade-long program that will support and accelerate development of the technology.

"We are on the cusp of a new era in automotive technology with enormous potential to save lives, reduce greenhouse gas emissions, and transform mobility for the American people," Secretary Foxx said in a statement.

Currently, each state has individual laws regarding the burgeoning technology. This forces automakers and early adopters like Google to juggle multiple sets of rules and regulations, depending on where in the country the technology is being built. For example, California recently decided that a human "driver" be present at all times should something go wrong. This program should reduce the number of hoops companies have to jump through by creating an overarching, national-level development framework.

DARPA is Developing an Unjammable Communications System

As reported by Engadget: Electronic warfare (EW) techniques -- from killing missiles with microwaves to downing drones with radio interference -- have become an integral part of modern wars. 

This issue is exasperated by the fact that both sides in a conflict must also compete for the finite spread of interference-free wireless spectrum with which they operate their multitude of wireless devices. 

To maintain American fighting superiority over the likes of China and Russia, both of whom are considered "near peers" in terms of EW combat capabilities, DARPA has developed an ultrafast chip to convert analog wireless signals into digital ones in record time.

DARPA's ADC, or analog-to-digital converter, will allow US forces to avoid being jammed by enemy EW methods by processing chunks of the electromagnetic spectrum about 10 times faster than what current-generation alternatives can muster. It takes over 60 billion analog and digital samples -- equivalent to roughly 1 Terabyte of data -- every second. This allows US forces to analyze more data from a larger swath of the spectrum in the same amount of time as current ADCs.

The chip isn't quite ready for the battlefield just yet, however. Its current iteration still draws far too much power to be useful in the field. DARPA is therefore partnering with GlobalFoundries to shrink the processor from its current 32nm fab down to 14nm. That should reduce its power consumption by 50 percent while maintaining its blistering performance. Next, DARPA will tackle managing the massive amounts of data this thing produces.