As reported by Quartz: GPS is a fine tool for finding a route between two places when there is a driver controlling a car, but as we near an era of fully autonomous vehicles, that technology is in need of a major upgrade.
That’s why Tesla is not just in the process of creating its own maps but is deciphering where each individual lane is on every road, across the globe. It’s doing this in part by tracking every one of its Model S cars each time a customer takes a drive, to learn where traffic typically moves. The project is immense but it is necessary if autonomous cars—which Tesla expects to be a reality in three years—are to work properly.
Tesla announced yesterday (Oct. 14) that it was upgrading all of its Model S vehicles with new functions that allow the cars to steer, switch lanes, and park autonomously (though it advised drivers not to actually take their hands off the wheel). Currently, tricks like autonomous lane-changing require a driver to indicate first, and for the car to rely on onboard sensors to decide when it is safe to change lanes.
But a fully autonomous car will need to know which lane it’s already in, how many lanes there are on the road, and where those lanes lead to—or end. Onboard sensors alone can’t generate this level of data. Elon Musk, Tesla’s CEO, believes this mass collection of data—Tesla plans to build around 50,000 cars this year—will set the company’s cars apart from rival auto makers who are preparing their own self-driving cars.
Certainly, tracking existing drivers will give Tesla a head start against “legacy” auto makers such as BMW, which, along with a consortium of other German brands, bought mapping software from Nokia earlier this year. But two more contenders for the autonomous car space may have an even greater advantage than Tesla: Apple and Google.
These companies could potentially track hundreds of millions of smartphones as their owners drive them from place to place, making their future road maps many times more accurate.
As reported by Mashable: Starting Thursday, yourTesla Model Swill have received an over-the-air software update, but not for free. The new 7.0 software will bring with it a new, more modern look to the operating system, some new instrument panel apps, climate control interface improvements, and — oh, yeah — Autopilot.
Autopilot, however, will cost every Tesla Model S owner $2,500, which probably qualifies it as the most expensive in-app purchase ever.
The Autopilot system is comprised of several different subsystems that include Autosteer, Auto Lane Change, Traffic-Aware Cruise Control, Side Collision Warning and Autopark.
The Autopilot suite of features that will be available to Model S cars built from September 2014 and beyond — essentially, any Tesla with the appropriate sensors. With the 7.0 update, drivers will notice the look of their instrument cluster has changed, displaying an update icon of the vehicle. This icon will show a detected vehicle ahead, lane marking readings, and also roadway obstacle warnings.
All exciting things — so let's dig into these new features one by one.
Autosteer
This is the new feature drivers will find most foreign and distinctive in the Autopilot suite, as the car effectively steers itself down the freeway. To activate Autosteer, the driver simply pulls the cruise control lever twice in quick succession. Turning on Autosteer also engages Traffic-Aware Cruise Control, which is Tesla's cruise control system that adjusts the speed of the Model S based upon the speed of the vehicle ahead.
How autopilot looks on the Tesla Model S dashboard.
IMAGE: NICK JAYNES/MASHABLE
Once Autosteer is on, a chime will ding. The sensors in the front of the Model S monitor the lane markings ahead and automatically adjust steering to keep the car within the markings. Now, if you were hoping that Autosteer would free up your hands for some idle work, like, say, whittling, I have some bad news. Like all other self-steering systems from other automakers, including Honda and Mercedes-Benz, the driver must keep his or her hands on the wheel at all times.
Tesla is quick to point out, though, that rainy, snowy, or foggy conditions will deteriorate the Autosteer performance. If Autosteer cannot read the lane markings, it will not be able to work. Autosteer will also cut out if the road the Model S is traveling on is too twisty.
In ideal conditions, however, the Model S will be able to keep itself within its lane without any input from the driver — accept the reassuring feeling that the driver's hands are still on the wheel.
If Autosteer can't confidently operate, it will display a warning for the driver in the instrument cluster. If the driver does not take over immediately, it will chime at you. If the driver still does not retake control of steering, the car will chime incessantly. If the driver does not take control after a few seconds, the Model S will bring itself to a swift stop in the lane and activate the emergency flashers.
Auto Lane Change
Now here's where the Tesla Autopilot suite gets interesting. When Autosteer and Traffic-Aware Cruise Control are engaged, the Model S can autonomously change lanes. This is the biggest distinction between Autopilot and any other carmaker's mild autonomy setups.
Say you want to pass the car ahead of you, simply signal in the direction in which you want to marge (which would be to your left because you don't want to pass illegally on the right, right?) and the Model S will move into the lane in the direction you indicated with your turn signal.
As it does this, the icon of the car in the instrument panel screen will go from a solid to a dashed line. Once the Model S determines the lane change is complete, the dashed line will return to a solid line and you can switch off the turn signal.
Side Collision Warning
Should a car or another object get too close to your Model S for its own comfort, it will alert you with visual cues on the digital instrument cluster. The graphics indicating an impending collision resemble radiating fluid lines coming off the side of the little digital Model S on the screen.
Arguably, this seems like a pretty innocuous extension of side blind-spot monitoring. Side Collision Warning does have an interesting feature, though. Should a car or object get too close to the Model S, it will autonomously steer out of the way while still keeping itself in its own lane. Intriguingly, this safety feature will activate even if the driver has disabled Side Collision Warning in the system settings.
Autopark
Should you you find a parallel parking spot that you don't feel quite up to the task of threading your Model S into, it can now do the duty for you. After you pass a suitable parking spot, a P logo will appear on the instrument cluster, indicating to you the car has seen the spot and is ready to park in it.
Put the vehicle into reverse and the Autopark guide will appear on the screen alongside the reverse camera image. Once you engage Autopark, the system controls both steering and vehicle speed. This is in stark contrast to a similar system from Ford, which relies on the driver to control vehicle speed with throttle and brake inputs.
Should you want to stop the Tesla Autopark system at anytime, you can do so by taking over steering operation, stepping on the brake or simply pushing the cancel button.
If you were hoping to whiz through town without a care in the world while your car searches for a parking spot, think again. Autopark will only recognize suitable parking spots when it is traveling below 15 mph.
Additional Updates
Of course, being Tesla, the 7.0 software update includes more than Autopilot. It also includes a new feature called Vehicle Hold, which is an extension to Hill Start Assist, which hold the car in place on steep inclines between the driver releasing the brake and applying the accelerator.
Vehicle Hold will also keep the Model S in place for extended periods of time. When engaged, the system is essentially a flexible parking brake, one that can be easily released by pressing the accelerator pedal.
Tesla has also updated the look of the instrument cluster energy usage display and clock in addition to the aforementioned vehicle icons that display the Autopilot functions.
As reported by Wired: NOW THAT THE Volkswagen diesel emissions scandal is a month old, the new VW Brand Board of Management is beginning to publicly discuss its plans for the future—and how they’ve changed. Step one: Use emission control systems that actually control emissions. Step two: Go electric.
As soon as possible, VW says, it will start equipping all its diesel cars in Europe and North America with AdBlue technology and selective catalytic reduction, a chemical process that breaks smoggy NOx down into nitrogen and water. It adds complication and expense—$5,000 to $8,000 per car—but it’s effective, eliminating 70 to 90 percent of NOx emissions.
VW’s big “advance” was the “clean diesel” technology that supposedly made this technology unnecessary on its smaller cars, like the Beetle, Jetta, and Audi A3 that are being recalled because they don’t meet emissions standards under real-world driving conditions.
Beyond diesels, VW announced it is “giving our product range and our core technologies a new focus,” lurching away from diesel and toward another way to meet increasingly strict CO2 and NOx emissions regulations in the US and Europe: an increased focus on plug-in hybrids and electric vehicles. One of those, the company announced this week, will be the next-generation Volkswagen Phaeton.
Details are scarce, but the press release is full of high-minded statements about the past and the future. The first Phaeton was an engineering marvel that turned out to be one of the finest executive sedans ever made. Too bad no one in America wanted a $70,000 Volkswagen.
Well, VW is trying again, but rather than making the world’s greatest gasoline-powered car, it’s looking at something more futuristic.
“The Volkswagen Phaeton has embodied the brand’s technological competence and brand ambition from the first generation onward,” the company says. It plans to make the new Phaeton “the flagship for the brand’s profile over the next decade.”
The new Phaeton has supposedly been in the works for a while. VW says it’s “redefined the current project,” though it’s not clear exactly what’s changing, and what kind of work needs to be done to get there. VW is calling for a Phaeton with “pure electric drive” with “long-distance capability.”
This is not an easy shift to make. To date, the pure electric vehicle market has been split in two, with the roughly $100,000, 265-mile range Tesla on one side, and more affordable, 85-mile range city cars on the other (like VW’s own $29,000 e-Golf). Without years of investment, it’s hard to believe VW can bring out a fully electric large sedan with serious range and an attainable price tag anytime soon.
That’s why VW says its “focus is on plug-in hybrids,” which combine an electric motor with an engine, and can drive powered either by electricity or by gasoline. VW’s sister brand Audi has already put plug-in hybrid versions of its R8 and A3 cars on the market, with the “e-tron” moniker. VW brands Bentley and Lamborghini are also planning to introduce plug-in hybrid versions of their upcoming luxury SUVs, so it’s not a stretch to imagine the tech coming out with a VW badge on the hood.
Along with electric drive, VW promises the Phaeton will deliver “an emotional design.”
Still, Volkswagen is a huge company with a ton of wicked smart engineers and designers (the diesel scandal aside) and we’re very interested to see what they come up with. But it may be too little, too late.
Spencer Dale, Governor Mark Carney, Jon Cunliffe and Andrew Bailey (L-R) arrive to release the Financial Stability Report, at the Bank of England in London.
As reported by Business Insider: Concerns about the climate change impact of burning the world's remaining oil resources mean the reserves will never be fully exploited, BP Chief Economist Spencer Dale said on Tuesday.
That assumption means the relative price of oil will not necessarily increase over time, Dale said at an economists' conference in London on Tuesday.
Bejing Air Polution
Burning existing reserves of oil, gas and coal would emit more than 2.8 trillion tonnes of climate-harming carbon emissions, much more than the 1 trillion threshold scientists have set to limit global warming to 2 degrees.
"Concerns about carbon emissions and climate change mean that it is increasingly unlikely that the world's reserves of oil will ever be exhausted," Dale said.
BP is among the world's top oil producing companies but it is also part of a group of energy firms that has called for the creation of a global carbon pricing mechanism that would limit investments in climate-harming forms of energy.
Technological advances will also reduce the cost to extract harder-to-reach resources, Dale said.
As reported by SlashGear: As arch-creator of concept car follies, Rinspeed is no stranger to outlandish designs, but just how practical isthe drone-equipped Etos? The self-driving concept can deploy a similarly self-flying unmanned aerial vehicle (UAV) from a bay at the back, with a variety of potential uses that include delivering forgotten items or filming particularly interesting stretches of road.
"For example, it could fetch a bouquet of flowers the driver ordered online and even deliver it directly to the lucky recipient," Rinspeed suggests. "It could also take a video of the wild ride on the favorite fun road and stream it live to friends!"
The Etos isn't going to get anywhere near production, used instead as a four-wheeled billboard of sorts for the customization projects that actually pay Rinspeed's bills.
Nonetheless, it isn't the first firm to suggest pairing a car with its own pet drone. Last year, for instance, Renault brought its Kwid concept to the Delhi Auto Expo, a chunky little urban SUV with a landing pad for a companion UAV on the roof.
Like with the Etos, the Kwid's drone could be used for taking aerial photography, but it might also have a role in spotting traffic jams and other road congestion up ahead, the French firm theorized.
Modern drones have no issues with basic following, and indeed many - like DJI's Phantom 3 - can be set to automatically track their operators. That usually involves a fixed altitude, or a preset route, however, rather than the sort of dynamic flight that an application like the car companies have in mind would demand.
It's when things get unpredictable, according to 3D Robotics - which launched its ownSolo smart drone earlier this year that promises easier GoPro shots courtesy of autopilot presets and a following mode - that the technology will need to raise its game.
"In short, yes, it is well within the realm of the possible to build a drone that can take off from a moving car, and then return to land, dock and recharge," company spokesperson Roger Sollenberger told me. "But as with any autonomous mobile technology, there are a number of problems to anticipate and solve before possible becomes practical. The anticipation is the important part, the unknown unknowns, etc."
Nonetheless, Sollenberger is confident that the tech side can mature sufficiently to meet the challenge. "[There] are some key technologies will preclude a lot of these issues," he explained, though whether they'll be practical is a separate question.
"Sense & avoid, so the drone will be able to navigate any obstacles in its path," Sollenberger lists, "the next generation of batteries, with more juice and a shorter recharging period; ground sensing (part of sense & avoid) so the drone would avoid things like [trying to follow a car through a tunnel]; air traffic control for drones, so that drones and other aircraft can 'deconflict' the airspace in real time as they fly; and of course, a regulatory framework - where and when can you launch from and land on your car? How fast can you go? - that can account for and incorporate new technology such as this."
"And this is the point where 'possible' and 'practical' meet, perhaps converge, perhaps not."
As for helping autonomous vehicles figure out what's happening on the route up ahead, even with their lofty perspective drones might not be the best source of information there.
Arguably more likely is a growth in car-to-car and car-to-X intercommunication, with autonomous and semi-autonomous vehicles effectively chatting between themselves and sharing real-time information on what hazards there might be around the corner. That data would have the benefit of tapping into the high-resolution radar, camera, IR, and other surveying data from the cars' own sensors.
It also saves you from worrying about mid-air drone collisions, or more mundane issues like running out of charge.
In short, while a flock of drones surrounding your car might make for a cool concept, for the moment the only eye-in-the-sky watching out for you is probably waiting to give you a speeding ticket.
As reported by Business Insider: If Elon Musk was not already running Tesla and SpaceX he might take up another big project: creating an electric jet.
In fact, when asked during a recent Marketplace interview what Musk would be doing if Tesla didn't exist, he revealed that he already has a design for such an aircraft ready to go.
“I do like the idea of an electric aircraft company. I think one could do a pretty cool supersonic, vertical, take-off and landing electric jet. That would be really fun... I have a design in mind for that," he said.
"Then of course there is the Hyperloop, I’d probably be working on that,” he casually added.
This is not the first time Musk has mentioned that he is interested in this kind of project.
During another recent interview with CNN Money, Musk said that the "ideal, long distance transportation mechanism is a supersonic, vertical take-off and landing electric jet."
Musk told Marketplace that excluding rockets, he believes one day all vehicles will be electric, so he thinks it's a promising venture.
“Aircraft and ships and all other forms of transport will go fully electric. Not half electric, but fully electric. No question,” he said.
Musk, though, said he has enough on his plate right now. The visionary billionaire already runs his electric car company and SpaceX and has big plans for both companies over the next few years.
One big goal of Musk's is getting Tesla's giant Gigafactory in Nevada into production. The company aims to make enough batteries to power 500,000 cars per year so it can finally begin production of its more affordable vehicle, the Tesla Model 3, which is expected to go into production in 2017 and be priced at $35,000.
"We have the Model X, that we obviously have to ramp up production of. We have a complete model of the Model 3, we have the battery pack business... and so there is a lot of work that needs to be done," Musk said. "But I feel at this point Tesla is in a really solid position. And at this point it’s a matter of how quickly can we bring an affordable electric car to market — how quickly can we bring affordable battery storage to market."
As reported by NBC News: There will be no arguing with the driver on this bus: the rides are free and there's no driver anyway.
Trikala, a rural town in northern Greece, has been chosen to test a driverless bus in real traffic conditions for the first time, part of a European project to revolutionize mass transport and wean its cities off oil dependency over the next 30 years.
Trials of the French-built CityMobil2 buses started last week and will last through late February.
A CityMobil2 driverless bus makes its way along a road in Trikala, Greece, on Tuesday.
Over the past year, CityMobil2 has been tried out near its base in La Rochelle in western France, on a campus in Lausanne, Switzerland, and near Helsinki, Finland, all in controlled conditions that produced no accidents.
But in Greece, a country of narrow, winding, hilly streets, stray dogs, bicycle riders and impatient drivers, the buses are up against real traffic.
The Greek government had to amend its laws to allow the testing and the city had to build a dedicated bus lane that deprived residents of downtown parking spaces.
The robot buses don't look like science fiction vehicles — more like golf cart meets ice-cream truck. Still, heads turn as the skinny, battery-powered buses hum through the streets. They seat only 10 people and are guided by GPS and supplementary sensors, including lasers and cameras, that send live data to a control center. The buses go no faster than 20 kph (12 1/2 mph), but the trials in Trikala (pronounced TREE-kah-lah) potentially represent a major advance for automated transport.
"There were cities bidding for this project all over Europe. They offered relatively restricted urban areas. But we said we could make it happen in a downtown environment and we won," said Odisseas Raptis, who heads the city's digital project department, e-Trikala. "We have a 2.4-kilometer (1.5-mile) route, the bus route. It's mixed with traffic, with pedestrians, with bicycles, with cars. ... That hasn't been done before."
Vasilis Karavidas, chief technician for the project in Greece, trained with Robosoft, the company that developed the bus, in the southwest French town of Bidart.
Although the driverless buses are fully automated with onboard navigation and obstacle detection systems, each vehicle will be monitored by a driver in the control center who can override the system, Karavidas said.
"It's as if they are in here and they can stop the bus if they want to, if something goes wrong," he said.
The buses are currently running without passengers, with full testing to start later this month when a fiber-optic network allowing faster data transmission is completed.
Six battery-powered vehicles will eventually be used in this farming town of 80,000 that has become hooked on high-tech.
Trikala already has already tested EU-funded pilot medical programs, including schemes to relay heart test data from home to the doctor's office and use tracker devices for Alzheimer patients.
A view out the front windshield of the driverless CityMobil2 bus. Thanasssis Stavrakis / AP
In the center of the city, a "digital tree" with solar panels allows benches to carry phone-charging outlets.
The 28-nation European Union is targeting gasoline use for city transport as one area where it wants to reduce carbon emissions. With oil prices and city populations expected to rise in the coming decades, a major shift to battery power and more shared transport could blur the line between private and public vehicles.
Senior transport analyst Philippe Crist at the International Transport Forum, an OECD think-tank based in Paris, says transport trends are hard to predict as the world moves more toward automation.
"We too often look at technological changes in isolation," said Crist. "There is a good chance that these technologies will create entirely new uses that we can only poorly grasp today. The reality is that everything is changing around these technologies and it is plausible that society may lose interest in owning cars or using fixed-service public transport — especially if these technologies allow better alternatives to emerge."
Models run by the think tank suggest that city transport could be made massively more efficient.
Crist said researchers looked at "shared and route-optimized on-call taxi-like services replacing all car and bus trips in a mid-sized European city. We found that these systems could deliver almost the same mobility as today but with 95 percent fewer vehicles."
Driverless cars and buses offer an easier way to optimize traffic flow while aiming to eliminate human error. That has transport developers working at both ends: adding automatic features to conventional vehicles while raising the bar for those that will have no driver at all.
So far, the CityMobile2 has had mixed reviews on the streets of Trikala. Not everyone is happy to lose parking spots or replace human jobs with machines.
Still, retiree Michalis Pantelis said he was proud that his city was selected for the testing.
"I think it's wonderful. Think how many people will come to Trikala to see this. It's new and innovative," he said, moments after a driverless bus passed by. "It reminds me of the toy cars my grandchildren play with."
This is not an easy shift to make. To date, the pure electric vehicle market has been split in two, with the roughly $100,000, 265-mile range Tesla on one side, and more affordable, 85-mile range city cars on the other (like VW’s own $29,000 e-Golf). Without years of investment, it’s hard to believe VW can bring out a fully electric large sedan with serious range and an attainable price tag anytime soon.
