Typically, one person drives the car while the other operates the EyeDrive system — an iPad equipped with the EyeDrive app, which provides routes and driving instructions. The vehicles must be driven in perfect weather conditions, and they must start capturing in the morning when the sun is at 30 degrees and drive until dusk when the sun returns to 30 degrees.

EyeDrive
(the software used in vehicles that gather data for Maps) pic.twitter.com/sRAXoCB8jA

— andrey (@YRH04E) August 3, 2020

According to 9to50Mac, the data is primarily being used for features like Flyover and Look Around. The vehicles are mostly seen in the US, but Apple is allegedly expanding them to Canada, Europe and Japan.

Apple Maps is still fighting its reputation as an internet punchline. In 2018, it promised to fix its Maps by using its own data, and we’ve seen it put in the necessary leg work. Since Apple finished rebuilding its Map data, it has rolled out useful features. It now shows COVID-19 testing sites and is using Map data to help public health officials. Plus, it now prioritizes grocery, food delivery and medical searches.

Thanks to the upgraded app, users will be able to set up their Neato D10 in just six steps. As promised, Neato vacuums now work with Siri Shortcuts, as well as Alexa and Google Assistant. You can map up to three floorplans, set “zone” cleaning, get cleaning summaries, choose turbo and eco modes and use remote start and scheduling.

Neato revealed the Neato D10, along with slightly paired down Neato D9 and Neato D8 models, at IFA 2020 in Berlin today. The Neato D9 and Neato D8 use the company’s LaserSmart mapping technology and have shorter run times than the D10, 120 minutes and 90 minutes, respectively.

The new product line is set to arrive in North America, Europe and Japan this fall, but the company has not yet revealed pricing. For reference, its existing Botvac D7 Connected cost $800. That puts Neato’s robotic vacuums in the range of the iRobot Roomba i7+, which originally sold for $1,000 but is now on sale for $700.

The base models would have ‘just’ two rear cameras and likely miss out on LiDAR and telephoto shooting, although they would still have OLED displays.

Other upgrades may be less dramatic but just as functional. The Face ID sensor array would not only get smaller, as previous rumors indicated, but work at wider angles — you’d have a better chance of unlocking your phone while it’s sitting on your desk. Big batteries might also be a staple of the lineup, with the 6.7-inch iPhone touting a battery over 4,400mAh (large by any standard) to handle both 5G and the 120Hz screen.

There’s no guarantee Apple will release these exact devices no matter how accurate the leaks may be. Last-minute changes or cancellations aren’t unheard of. If these devices do show up as depicted here, though, they’ll represent big strides forward in key areas, even if some of them amount to catching up with competitors.

The data was collected over the span of one year, and it comes from multiple self-driving research vehicles. It includes LiDAR and camera sensor data, GPS and trajectory information, as well as 3D point cloud and ground reflectivity maps. Ford is even sharing a plug-in with data visualization tools.

Ford

While other companies, like Waymo, have shared self-driving vehicle data in the past, the breadth and depth of Ford’s data is unusual. Because it was collected over an entire year, it includes a variety of weather conditions, including rain, sun, clouds and snow. The data was gathered in the Metro Detroit area, so the vehicles experienced dense urban areas, freeways, tunnels, residential neighborhoods, airports, construction zones and pedestrian activity. That should give research access to the kinds of diverse scenarios self-driving vehicles will find themselves in, Ford says.

Plus, while most datasets only offer data from a single vehicle, this data comes from multiple vehicles. That means researchers can explore, for instance, what happens when two vehicles encounter each other. One might be able to detect things that the other cannot “see,” which could lead to development around multi-vehicle communication, perception and path planning.

The data is available through a collaboration with Amazon’s open data program, and more info can be found at avdata.ford.com. The first set of data is available now, and Ford says it will continue uploading the rest of the data logs.

Ford has tested its self-driving vehicles and delivery services in a handful of cities — though it clarified that this data is separate from the work it’s doing with Argo AI to develop a production-ready system. Ford still plans to launch its self-driving car fleet in 2021, but last spring, CEO Jim Hackett scaled back those ambitions. He said the fleet’s “applications will be narrow, what we call geo-fenced, because the problem is so complex.”

More like a laptop than ever

When I reviewed the third-generation iPad a year and change ago, I half-jokingly wrote this:

“I know there’s zero chance of Apple listening to me, but I have to try. Hey, guys, can you please figure out how to make a trackpad work on this thing? I work in text all day, and it would make my life so much easier.”

Well, about that. After introducing support for mice and trackpads as a somewhat limited accessibility feature last year, Apple embraced them more fully in the new iPadOS 13.4 update. Once that’s installed, every iPad that runs iPadOS — even non-Pro models — can benefit from this expanded flexibility.

That new, $300 Magic Keyboard with its integrated trackpad won’t be available for another month and a half at the least, and in the meantime, there’s nothing else quite like it to use instead. (For the record, Logitech recently showed off cases with keyboards and built-in trackpads, but those aren’t meant for the 11-inch or 12.9-inch iPad Pro.) This means that, to get the most out of this new software, you’ll have to use something like Apple’s Magic Trackpad, and I’m hardly thrilled at the idea of carrying one around for when I need to be productive. Then again, I guess I’m not going anywhere soon: Everyone in New York has compelled by the state government to stay at home to help contain the spread of the novel coronavirus. That means no meetings or briefings where I need to quickly jot down notes, no idle writing sessions at local coffee shops. But I digress.

After using this new version of iPadOS, it’s clear Apple has done something pretty remarkable here: It made trackpad support feel natural. Getting started is easy: Turn on the trackpad, connect to it from the iPad’s settings and, uh, that’s it. It’s not quite Apple Pencil-easy, but it’s close. Once that’s done, you can fiddle in the settings menu with the cursor speed, scroll direction and tap-to-click.

The best thing about the experience is that it works almost exactly the way you’d expect it to. You move your finger around, click things and they respond. That may sound basic, but the sheer efficiency over reaching up to touch the screen is tremendous. And as you whip that tiny, circular cursor around the screen, you’ll see it change shape: It’ll become a classic I-bar when you hover over text, making it much easier to select exactly what you need. And when you get close to one of iPadOS’s interface elements, like the back buttons or the Control Center shortcut, the cursor sort of latches onto it. It’s a simple, elegant solution, as are the gestures.

A three-finger swipe up takes you to the home screen. A three-finger swipe to the left and right cycles you through your open apps. A two-finger tap acts as a right-click. There: You have everything you need to get going. It helps that, despite performing somewhat different tasks, all of these gestures are supported in macOS too, so I didn’t have much trouble getting comfortable. There is one gesture that continues to throw me, though. When your iPad is propped up and you want to unlock it with the trackpad, you move the cursor to the bottom of the screen and continue pushing down. It feels much different than the “Swipe up to unlock” I’m used to, and it’ll take time before my muscle memory adjusts.

I’ve mostly been thrilled with how this feature turned out, and I’m starting to think I might have to carry a Magic Trackpad around full-time after all. That doesn’t mean everything is perfectly peachy here, though. While most apps I’ve tried work fine with a trackpad, I did run into some notable hiccups. When exporting the photos for this review, for instance, I noticed that one of Lightroom’s menu options wouldn’t respond to a tap on the trackpad at all; I had to reach up to the screen instead. Honestly, it wasn’t much of an inconvenience, but I’m sure there are other apps out there with similar issues. More problematic was that Google’s suite of productivity apps — which we lean on extensively for editing around here — didn’t play well with iPadOS’s trackpad text selection.

Unfortunately, mouse support is considerably more limited. I’ve been testing this iPad with a Bluetooth mouse and a Magic Trackpad, and while the basics work much the same way, you lose access to all of those handy multitasking and navigation gestures. (I don’t have one of Apple’s touch-sensitive Magic Mice, but a few tweets suggest they don’t work with those gestures either.) Even though this update makes the iPad more flexible, it’s still a touch-first device. The difference is, now you don’t have to touch it directly.

These changes are a huge deal, if for no other reason than proving that Apple is serious about giving its users more flexibility in how they use their devices. It won’t, however, give you everything you need to give up on your trusty laptop. For that to happen, Apple would have to dramatically reshape the iPadOS experience, from rethinking window management (you can still only run two apps at a time in Split View) to making its arcane multitasking gestures a bit more obvious. Had you asked me when the last big iPad refresh happened if Apple would pour its resources into an endeavor like that, I’d have been pretty pessimistic. Now, it seems all but inevitable — I’m especially curious to see what Apple has to say about iPadOS’s new path at WWDC. Unless it gets canceled like Google I/O, that is.

The Precept also focuses heavily on eco-friendly design. The seat surfaces, bolsters, headrests and carpets are made out of recycled materials like 3D-knitted plastic bottles, cork vinyl and even recovered fishing nets. The interior panels and seatbacks, meanwhile, are made from flax-based composites that are not only kinder to the planet (80 percent less plastic waste), but lighter — they weigh half as much as conventional panels.

And yes, there’s plenty of technology on the inside. The Precept has a 12.5-inch driver display as well as a vertical 15-inch center touchscreen, and it uses a “next-generation” Android interface with improved Google Assistant voice control and the ability to watch videos while you’re charging. More importantly, it knows where your attention is. The interface uses eye tracking to adjust the interface based on where you’re looking, and proximity sensors both show more info and enlarge controls when you bring your hand close to the touchscreen.

The Precept will be on display at the Geneva International Motor Show starting on March 5th. You probably won’t see this exact car on the road, at least not in every country. US regulators aren’t keen on rear-view cameras, for example, while the rear view might also raise some concerns. And like most concepts, Polestar might have to tone down the design or features to make it road-ready. This is a fairly practical concept, though, and we wouldn’t be shocked if something similar hits the streets.

LiDAR is the final piece of the sensor puzzle for Bosch’s self-driving ambitions. The company said it designed the sensors for all autonomous driving use cases.

Laser-based LiDAR systems piece together a high-res, real-time 3D map of the surrounding area, which car systems can use to better understand everything happening nearby and react accordingly. The sensors work in concert with radar and cameras to help cars navigate with optimal safety.

Bosch isn’t the only company working on LiDAR sensors, but it’s a major supplier. If it can keep a tight lid on production costs, it could help make semi-autonomous and fully autonomous cars more affordable.

Last year, Luminar unveiled a LiDAR system that’s said to cost less than $1,000 for production cars. Waymo also said last year it’d start selling LiDAR sensors, possibly for less than $5,000 each. Those are significantly cheaper than some other sensors, which have cost up to $75,000 over the last few years. It’s not clear when Bosch will start selling its own sensors or how much they’ll cost, but at the very least more competition should help drive down LiDAR prices.

Meanwhile, Bosch said it recently implemented artificial intelligence in its self-driving camera tech. It can detect and categorize objects and measure their movement. It can also recognize “partially obscured or crossing vehicles, pedestrians and cyclists quickly and reliably.” This, Bosch suggests, should help driver assistance systems trigger a warning or enable emergency brakes more effectively if needed.

Everyone’s AirPods will die. We’ve got the trick to replacing them.
Geoffrey A. Fowler,
The Washington Post

If you hopped on the AirPod bandwagon early, your earbuds are likely near the end of their life. The Washington Post details what happens when your true wireless audio accessory stops working, including the fact that “battery service” really just means giving you a new pair and tossing your old ones in the recycling bin.

How Space Archaeology Works

Whenever archaeologists apply any form of air- or space-based data to the assessment of modern landscapes, attempting to locate long-buried rivers or hidden ancient sites, they are doing “space archaeology,” also called “satellite archaeology” or “satellite remote sensing.” NASA shoulders the ultimate name blame. In 2008, NASA began its “Space Archaeology” program,11 funding scientists to apply satellite data sets to large-scale archaeological research projects. If NASA calls what I do space archaeology, who am I to disagree?

Interpreting satellite imagery is part science and part art. All remote-sensing specialists must start by learning the language of light, and it is not easy: what appears as a simple high-resolution photograph on your computer screen is so much more. Each pixel on the image is representative of an exact area on the ground. The light composing the pixel represents not only the visible part of the light spectrum, but the near, middle, and far infrared, depending on the satellite-imaging system. Additionally, everything on the Earth’s surface has its own distinct chemical signature that affects the light it reflects: much as we all have distinct signatures when we write our names, different materials show up uniquely in the light spectrum.

For example, sand appears very different from forest on the satellite imagery. We can see this easily with our own eyes. When you need to discern different tree species within the forest, this is where chemical signatures come into play. A group of oak trees emits a different chemical signature than does a group of pine trees. Visually, they might appear as the same green to us, but using different parts of the infrared spectrum to visualize subtle vegetation health differences, we can perceive color variation.

Remote-sensing specialists can exaggerate these differences by assigning “false color” to the images, to highlight individual classes of surface features. Within remote-sensing programs (like Photoshop color replacement with an attitude), you can choose any color for any cluster of pixels. While it’s recommended that users choose classes closely resembling their real-life counterparts—for example, green tones for vegetation, gray for buildings, brown for soils—you can choose any colors you want. Satellite images shown at conferences or in publications sometimes look like bad acid trips.

Scientists shop for specific types of satellite images to suit the data they need. Each satellite is different, and there are over 1,700 of them up there. Most are lower-resolution weather or large-scale satellites, with resolutions of 15 to 30 meters. These are the images most used, not just because they are free, but because there are millions of images going all the way back to 1972 that highlight short- and long-term landscape changes. In addition to these free images, there are high-resolution images recorded by sensors such as DigitalGlobe’s WorldView-3 and -4 satellites, with resolutions of between .31 and 1 meter, where a single pixel represents an area between the size of an iPad and a bodyboard. Everyone looking at satellite imagery extracts pixel-based data to detect subtle short-term versus long-term changes, or to detect features. We tweak and test algorithms depending on our research questions, and eventually, through sheer dumb luck or a moment of genius, we find something of interest, usually because we’re scraping the barrel bottom of possible techniques. When it turns out to be dried snot on our computer screen, this being science, we go back to the drawing board and try again.

It Isn’t All “Aha” Moments, Except When It Is

People think that remote-sensing work is all about the “Aha” moment, the moment when a single click of a button reveals secrets hidden in plain sight. It isn’t. A typical remote-sensing specialist will spend dozens of hours per week in front of a computer screen, often cursing due to program crashes. When something does work, there is additional swearing, because you have forgotten to record the exact steps you took to reach that point. And you must start over. It’s about learning, about refining the process.

Then again, “Aha” moments do happen. One of my favorite remote-sensing stories unfolded at the well-known Maya site of Caracol in Belize, which dates back over 1,000 years. In 2008, a new laser-imaging technology called LIDAR, for LIght Detection And Ranging, was just warming up at the starting lines.

Diane and Arlen Chase, a gregarious and generous archaeologist couple at the University of Nevada, Las Vegas, had worked at the site for nearly 30 years. When a keen biologist, John Weishampel, of the University of Central Florida, first asked the Chases about using LIDAR at Caracol, they told him that they were skeptical. They had never heard of it, but they were understandably enthusiastic about the idea of bringing more funding to their site. After decades of toil, they almost hoped they hadn’t missed anything major.

They told him to go ahead with his grant application—he could try and peer beneath dense rainforest canopies using LIDAR if he wanted. It sounded like fun and wouldn’t do anyone any harm. John, now grant in hand, commissioned an airplane from the United States to collect the point cloud data, or hundreds of thousands of points from the top of the vegetation down to the forest floor, in a large area surrounding the site.21 If you were to look at the area on Google Earth, all you would see is rainforest—a sea of green, with
nothing suggesting anything ancient, aside from a few well-known limestone pyramids peeking through the tops of the trees.

After he had processed all the data, John displayed the images for Arlen and a small group. Arlen’s exact words were: “Holy shit!” The same thing was on everyone’s minds. Another astonished colleague said that this was the data to launch a hundred PhD dissertations.

The next day, Diane called John: “Arlen’s been stuck to his screen all night looking at the images. And he’s missed dinner and breakfast.” In a single night, the entire field of Mesoamerican archaeology had changed permanently: Arlen had found more ancient Maya sites than he had in 30 years of combing the jungle. Today, he can find 500 new Maya features before lunch from his desk in Las Vegas.

Such wholesale rethinking is not the product of a single flash of technical brilliance, but rather the result of decades of often serendipitous developments in the field of archaeology. To understand this takes a brief nosedive into the history of seeing ancient sites from afar.

Excerpted from Archaeology from Space: How the Future Shapes Our Past by Dr. Sarah Parcak (Henry Holt & Co, 2019). All rights reserved.

It’s essentially like looking into the darkness of the universe in its grand expanse of nothingness while contemplating the lifeless hole that has engulfed your soul as you age towards oblivion but with a BMW badge, grille, and headlights. It really is something you have to see for yourself because the car loses all dimension and looks like a silhouette even when the lights in the Coachella-rave-tent inspired room are turned on.

But even blackest black can’t escape the power of lasers. LiDAR outfit Ouster took one of its sensors into the room to see what they could see and caught the vehicle in all its X6 darker-than dark-glory.

The sensor relies on a reflected laser and ambient light to “see” objects, and when you have something this black there’s very little reflectivity. Mark Frichtl, Ouster co-founder and CTO explained his theory why a pigment that just sucks up light was caught on the company’s LiDAR system.

Covering the X6 in Vantablack VBx2 reduces both (laser and Ambient light) of these parameters dramatically. At one-percent total hemispherical reflectance, it’s 10-times less reflective than a typical ‘low-reflectivity’ scenario for specifying maximum ranges for LiDAR.

The Ouster OS1-64 LiDAR sensor is specced at a maximum range of 40 meters for a 10 percent reflective target, so there’s a maximum range of around 22 meters for a target that’s one-percent reflective, like VBx2-coated X6.

In other words, it sees the car, but it needs to be close to it to do so. Fortunately, some parts of the car are very reflective so even at long distances, the Vantablack X6 is visible.

“You’ll also notice that certain parts of the X6 are designed to be extremely reflective, like the license plate and retroreflectors in the headlights and signals. Those parts would be easily picked up at more than 100 meters away. And you can see that in the video, with the license plate being one of the brightest objects in the scene,” Frichtl told Engadget.

Of course, BMW will never sell a car with Vantablack. But if someone decides to spend all their money on a custom paint job, the vehicle will still be detectable by sensors as long as they don’t cover the entire vehicle (including the windshield).

LiDAR has become one of the go-to sensors for most automakers to help vehicles see the road, especially for driver’s assistance features. In addition to adding redundancy systems that already have cameras and Radar, LiDAR’s ability to create high definition images from the bouncing of light off surfaces from far distances gives vehicles a more robust piece of data about an object.

This demonstration is further evidence that a suite of sensors on vehicles is always better than just one.