Urban Jungle a Tough Challenge for Google’s Autonomous Cars

As reported by MIT Technology Review: After catching the world and the auto industry by surprise with its progress with self-driving cars, Google has begun the latest, most difficult phase of its project – making the vehicles smart enough to handle the chaos of city streets.

But while the company describes its work with its typical tight-lipped optimism, academic experts in robotics are cautious about the prospects of fully autonomous vehicles. They estimate it will be decades until they can perform as well as human drivers in all situations – if they ever do at all.

Google’s cars make extensive use of detailed maps that describe not only roads and restrictions such as speed limits, but the 3-D location of stop lights and curbstones to within inches. The company is now working to make its vehicles capable of seeing and understanding the kind of unexpected obstacles that don’t appear on those maps and are particularly common in urban areas, said Chris Urmson, the director of the project, last week.

“Obviously, the world doesn't stay the same,” said Urmson, speaking at a conference bringing together academics and auto-industry engineers working on autonomous driving. “You need to be able to deal with things like temporary construction, and so we've been putting a lot of effort into understanding the semantic meaning of the world.”

For example, an autonomous car should be capable of recognizing that a school bus is different from other vehicles of a similar size and may behave differently, said Urmson.

Urmson showed video of a prototype Google car navigating through a real-life construction zone marked by flashing yellow arrow signs, and even stopping when a “construction worker”—actually a Google employee—waved a hand-held stop sign.

Having cars understand those types of hazards is crucial to Google because of a recent change in the direction of its project. The company’s original prototypes were based on conventional vehicles, and a human passenger could use the steering wheel and brake pedal to intervene in the event of a glitch. But in May, Google said that humans couldn't be counted on to stay focused enough on the road (see “Lazy Humans Shaped Google’s Autonomous Car”). It unveiled a new prototype without a steering wheel or pedals and said research would now focus on making vehicles that are 100 percent autonomous—leaving no room for error.

Academic experts at the conference say Google is taking on some of the hardest problems in artificial intelligence and robotics, essentially trying to replicate the ability of humans to effortlessly make sense of their environment. That’s because driving safely relies on much more than just knowing to avoid big objects, such as people or other cars, or being able to recognize symbols such as a stop sign.

Humans make use of myriad “social cues” while on the road, such as establishing eye contact or making inferences about how a driver will behave based on the car’s make and model, Alberto Broggi, a researcher at Italy’s Universita di Parma, told MIT Technology Review.

Even if a computer system can recognize something, understanding the context that gives it meaning is much more difficult, said Broggi, who has directed several major European Research Council grants in autonomous driving. For example, a fully autonomous car would need to understand that someone waving his arms by the side of the road is actually a policeman trying to stop traffic.

When surveyed by the conference organizers, the 500 experts in attendance were not optimistic such problems would be solved soon. Asked when they would trust a fully robotic car to take their children to school, more than half said 2030 at the very earliest. A fifth said not until 2040, and roughly one in 10 said “never.”

Several of them told MIT Technology Review they wouldn't be surprised if self-driving cars were, for many decades, limited to specific, well-controlled settings, such as construction sites and campus-like environments with low speed limits and minimal traffic.

Most big auto companies are exploring self-driving cars. One of them, Nissan, caused a stir last year when it predicted it would be selling them by 2020. Last week, though, Nissan used the conference to dial back that forecast, saying instead that cars by the end of the decade will be able to handle selected tasks, such as parking and freeway driving. Despite being bullish about its technology, Google doesn’t make predictions about when fully autonomous vehicles might arrive.

John Leonard, an MIT expert in autonomous driving who attended the conference, says that he and other academics find themselves constantly battling the assumption that all of the technology challenges associated with robotic cars have been solved, with only regulatory and legal issues remaining. “It’s hard to convey to the public how hard this is,” he says.

Leonard stands by a comment that earned him some online criticism in an MIT Technology Review story last year, when he predicted that he wouldn’t see a self-driving Manhattan taxi in his lifetime (see “Driverless Cars Are Further Away Than You Think”).

The Giants Are Putting the G in GPS

As reported by the Wall Street Journal: The culture around Giants football is about as new or modern as their head coach, Tom Coughlin.

Their stadium and practice facility are younger than most, but both are filled with memorabilia and banners dating back to the team's inception in 1925. Likewise, Coughlin—the NFL's oldest coach at age 67—began his career back in 1970, before anyone was too concerned with hydration or soft-tissue injuries.

But to watch the Giants practice now, with their heat breaks and global-positioning devices, you might accuse Coughlin of being forward-thinking—perhaps even cutting edge.

"Coughlin is definitely on the forefront of the GPS," right tackle Justin Pugh said Sunday.

The team has outfitted its players with GPS devices strapped to their backs under their uniforms. The technology can provide the team with information such as top speed, burst and how long the player has run in terms of time and miles.

"We're getting all kinds of information," Coughlin said after Sunday's practice. "It's player related so we have information coming every day on every player."

The Giants aren't eager to discuss the specifics of that information. The team declined a request to make someone available who could speak knowledgeably about the program, which is now in its second year.

The players, on the other hand, are willing to talk about the GPS devices and other gadgets aimed at measuring recovery and preventing injuries.

There are those, such as defensive tackle Cullen Jenkins, who admittedly give the training staff a hard time. Jenkins likened the experience of strapping a GPS device on his back to being a "test subject." But at this time of year, when jobs are won or lost over health concerns, many players are eager to digest the data.

"I've never done the GPS," said left tackle Charles Brown, who signed with the Giants as a free agent after four years in New Orleans. "I like it, though. It's a way to monitor everything we're doing…heart rate, how much running you've done and at what speed and what angles."

The information gleaned from the GPS can be used in any number of ways.

As Coughlin explained last week, the data can help the coaches structure practices so they know when to do extra stretching or potentially cut down on reps. And with the temperatures soaring during training camp, the Giants have been taking breaks during practice to get out of the sun and hydrate.

Some of the players are even using the information as a means of competition.

"We always look around: 'Who had the most burst? Who had the highest top speed? The workload?'" explained Pugh. "Obviously, the more reps you get, the higher your workload is gonna be. And it definitely helps. The one day we had the recovery day and regeneration, and it helped us with the next practice the next day.

"I always have the top mile per hour," he continued, "so I always hold that over everybody."

Pugh even spoke of a watch that can measure players' sleeping habits. The team only has "like 10," he said, so not everyone has tried it, but the overall purpose is clear: The Giants want to know how much players are working and how they're recovering afterward.

Kicker Josh Brown doesn't wear the GPS or the watch, but he has spent years studying the effects of heart rate and recovery on his own game. As a rookie in Seattle back in 2003, Brown said, he kicked around "50 balls a day." Now he might attempt as few as 18 kicks, as long as they all "look exactly the same," to avoid fatigue and muscle issues.

The Giants are still mastering the GPS technology.

Back in May, after drafting LSU wideout Odell Beckham Jr. in the first round of the NFL Draft, general manager Jerry Reese boasted about the intensity with which the Louisiana native practiced.

"All of the teams use the GPS like we use, and early in camp, they said his GPS registered from Baton Rouge to New Orleans," Reese said. "That is how hard he works. Because he does all of the special-teams stuff as well, he gets a lot of mileage on his GPS."

Perhaps that mileage could have served as a warning. Beckham has been battling hamstring issues since the spring, and while the direct cause of the injury is unknown, the team will likely look at his workload through a different lens going forward.

Vehicle Technology Disruptors: BMW Vs. Tesla

As reported by TechCrunch: Jill Lepore generated quite a fracas in Silicon Valley with her New Yorker article that questions disruptive innovation and posits that large incumbent companies often survive and subsume disruptive technology with small incremental gains. Fortunately, we have a live Petri dish: BMW’s new electric i3 is an ongoing case study of a legacy manufacturer facing an innovator’s dilemma in the face of Tesla, a very aggressive new competitor with next-generation technology.

Elon Musk has defined the standard for a future mass-produced electric car – it must cost around $40,000, have a range of 200 miles, and be comparable to a BMW 3 series. In order to achieve that audacious goal, Tesla is embarking on a plan to build a “Gigafactory” capable of producing batteries at an efficient and lower cost that would make such a dream car feasible. Investors are betting that Tesla will be able to dominate the electric car market when it achieves scale, continuing a growth rate that values Tesla at $28 billion even though it only produces 35,000 vehicles a year. It is interesting that Musk directly compared the Tesla’s upcoming mass market Model 3 directly to the BMW 3 series, given that BMW is now delivering its new i3 to the US market in accessible volumes.

There are lots of great lessons for entrepreneurs to learn from watching the BMW versus Tesla battle since cars are so tangible and manufacturer sales tactics are so transparent.

Brand

Even though it has a “3” in its name, the i3 is decidedly not a 3 series BMW. It is two feet shorter, and should instead be in the BMW 1 series product family. The i3’s electric range of 80-100 miles makes it more similar to electric cars like the Nissan Leaf and the Chevrolet Volt and nowhere close to a technological wonder like the Tesla Model S.

Despite its limitations, the i3 is clearly resonating, with rave reviews and a price that is spiking over the last month on TrueCar, indicating high initial demand in the United States. BMW has (mis)used the power of the 3 series brand to its benefit, and can now add features like longer length and range incrementally as battery technology improves.

Lesson: Legacy companies can mislabel their products to leverage their brand, especially if an upstart compares itself directly to a particular model.

Technology

BMW invested tremendous resources in its electric car platform to develop an all-electric vehicle platform, and it is willing to integrate legacy technology in order to deliver immediate value to its customers. Conversely, Mercedes chose a partnership route and is buying the drivetrain and battery technology for its upcoming electric car from Tesla. Both BMW and Mercedes are well ahead of Tesla in advanced vehicle technology like self-parking and cruise control that can automatically follow highway lanes and maintain distance from other vehicles.

Rather than waiting for battery technology to evolve to make an all-electric car with a 200-mile range at a mid-range price point, BMW is selling an optional “range extender” consisting of a two cylinder motorcycle engine that maintains the batteries at a 5 percent power level and extends the car’s range an additional 80 miles. Since the range extender powers the batteries rather than a gas engine, the i3 is not a hybrid, but the range extender can be continually refilled so that the car is never stuck without power. It’s a total hack, but is well thought out and competitive. BMW’s engineers must have been giggling when they came up with this one.

With the i3, BMW has delivered a “good enough” luxury electric car for the urban driver and average commuter, who can also optionally use the car for longer trips without having to plan for supercharger stations.

Lesson: Legacy companies are often willing to hodgepodge new technology with their older technology to stave off new competitors.

Volume 

Tesla shipped its first car in 2006 and is expecting sell 35,000 Model S sedans in 2014, or roughly 17,500 units in the second half of 2014. BMW started selling i3’s in 2014 and sold 6,000 i3’s in the first half of the year, primarily in the European market, which now has 3 to 6 months waits for the car. Now that demand is spiking, BMW increased production to 20,000 units annually and is now producing 100 units a day, a run rate of over 30,000 units annually. The fact that a legacy manufacturer is on the verge of outselling Tesla in its own luxury electric segment in the first year of shipping is fascinating given Tesla’s superior product and years of market lead.

Lesson: Once legacy companies have hodgepodged technology, they can produce it at scale.

Sales

While Tesla is right in attempting to disrupt the antiquated dealership business model, BMW will be able to leverage its extensive dealer network to deliver to consumers worldwide, and consumers can use web services like TrueCar and Beepi to bypass the hassle of negotiating with dealers on the price of a new car and trade-in amount. BMW also has access to a deep well of financial incentives to drive consumers to buy cars. Auto manufacturers and their dealers are fighting Tesla with regulatory measures to slow the company down and limit market penetration.

Lesson: Innovators should not underestimate the power of a legacy company’s large, lumbering sales channel.

Market Entry

Tesla had to enter the market at the high-end in order to deliver batteries capable of long ranges at a margin that would deliver profits to fickle investors, years before it could deliver a mass market mid-range vehicle. BMW’s breadth enables it to enter at the mid-market and then move up into the ultra high-end next year in the U.S. with its i8 supercar. To BMW, the distinctive, urban-friendly i3 is essentially a rolling advertisement for BMW’s innovative and green future, so the company could even sell them at a loss and come out ahead.

Tesla’s high-end first approach could turn into a liability as the Tesla S is quite large and therefore not well suited for urban environments – it is wider than and almost as long as a 7 series BMW. Large, luxury four door sedans are typically purchased by upper middle class men over the age of 35 which, as a member of the demographic, I unfortunately have to admit we’re not exactly the most hip crowd. BMW examined the market thoroughly and is targeting hip, young, urban professionals with the i3’s forward design, a smaller urban-friendly size, and the brand’s proven appeal with a younger demographic.

Lesson: Legacy companies are often in numerous segments of a market and leverage their scale to beat an upstart’s roadmap.

Who Exactly Is Getting Disrupted?

The big question is what industry exactly are electric cars are disrupting? At first it seemed like the legacy auto manufacturers would not be able to step up to an electric car challenge. They have widely adopted hybrid electric cars, are now delivering somewhat competitive electric cars, and are continually experimenting with hydrogen fuel cells. From a broader view, it is possible that ExxonMobil and Chevron will be more disrupted by electric vehicles rather than BMW and Chevrolet.

Elon Musk is an entrepreneurial hero who is concurrently disrupting the passenger vehicle, space transportation and electric utility industries. Some of the legacy companies in those industries were bound to wake up at some point and respond aggressively. Fortunately, Musk can rest assured that the United Launch Alliance will not be as agile against SpaceX as BMW has been against Tesla!

Three Car Technologies That Know When You're Too Sleepy to Drive

The Internet of things won't just change your home life.  It will also affect the way you drive, by keeping you alert and
preventing you from causing accidents.

As reported by GigaOm: Driverless cars like the one above, are still quite a way off into the future, but it’s easy to understand why robots might be safer drivers than people. We do dumb things like texting while driving, driving drunk and even falling asleep behind the wheel. So it’s interesting to think about how connected cars and the quantified self might meet in ways that help keep bad behaviors off the roads.

One way to do this is to put an ECG sensor in the seat of a car, so it can can measure a driver’s heart rate to detect how alert he or she is. Researchers at the Nottingham Trent University in the U.K., backed by the Technology Strategy Board, aim to figure out how to parse the data and separate the useful bits from the noise. They also want to determine the best way to integrate sensors into the fabric of a car’s seat. If the seat detects a sleepy driver, it would alert him or her, or possibly take control of the vehicle if the person doesn't respond. It’s like a fancier lane-drift detection system.

Another similar project (hat tip GizMag), called HARKEN (heart and respiration in-car embedded non-intrusive sensors) is a European effort to implant sensors that monitor respiration and heart rate into a car’s seat and seat-belt. The sensors detect when a driver falls asleep behind the wheel. The two-year project ended in June with a working prototype and will be tested on closed roads soon.

Finally, there’s Nissan’s concept project, Nismo. It includes a watch that tracks a driver’s biometrics and integrates with the car to share information about heart rate and road conditions. The concept comes from Nissan’s racing and high-performance division, so this is less about tracking when the driver is falling asleep and more about keeping a race car driver at the peak of performance. However, the idea that such integration might one day be available to the average consumer via in-car devices like Automatic or Zubie, connected to a fitness tracker such as the Withings Pulse or Jawbone.

In fact, with Automatic and Jawbone (and Withings) channels on IFTTT, maybe there’s a amateur-hacker solution to be made sooner rather than later. All we need are the right triggers.

Australian Team Smashes 26-Year-Old Electric Car World Record

As reported by C/NET: Sunswift, a solar car racing team from the University of New South Wales, has today broken an electric car world record that has stood since 1988. The record, overseen by the Federation Internationale de l'Automobile (FIA), measures the average speed of an electric vehicle over a 500km distance on a single charge.

The record-breaking vehicle is Sunswift's fifth car, the Sunswift eVe. The vehicle previously ran in the 2013 World Solar Challenge, a 3,000km solar car race in Australia that runs from Darwin to Adelaide. The car has a top speed of 140kph with an electric-only range of approximately 500km, or up to 800km when its solar cells are also active -- all while using about as much power as a kitchen toaster when travelling at freeway speeds. The solar system was turned off to adhere to the electric car specific record attempt.

Late Wednesday afternoon from the track in Victoria, Australia, they tweeted to confirm their success.

IT'S OFFICIAL: We broke the record!!!!!!
— Sunswift (@sunswift) July 23, 2014

Sunswift smashed the existing record of 73 kilometres per hour, achieving a final new record of over 100kph with a final official result awaiting confirmation with the FIA.
For the record attempt the team of university students was required to work with FIA accredited professional drivers. One of the drivers, Garth Walden, was particularly pleased with the opportunity.

"As a racing driver you always want to be on the podium and it's not everyday you get to break a world record," said Walden. "I really enjoyed hanging out with the team and being part of history."

For the Sunswift team, the result was about proving the future of everyday viability for solar electric vehicles -- achieving a speed that shows an electric vehicle can sustain freeway speeds over an extended range.

"Five hundred kilometres is pretty much as far as a normal person would want to drive in a single day," said Hayden Smith, project director and engineering student.

"This is really about curing people's fear of the lack of speed and curing their range anxiety, showing this car can travel at high speeds for long distances which is really what everyone wants."

The Sunswift team is now working on modifying the car to make it street legal and expects to be able to drive it on public roads in 2015.

UNSW's Sunswift eVe breaks a 26-year-old FIA electric car speed record. Dave Cheng/CNET

What it's like to drive eVe
We had a chance to take eVe for a spin in the weeks leading up to the record attempt at a track in Sydney. The car is built like a race car, made for efficiency and not for comfort, with a low-slung racing seat and incredibly stiff suspension.

Acceleration and regenerative braking were both managed via paddles on the steering wheel instead of foot pedals, with only the mechanical brake at your feet. It was readily apparent the vehicle would need a lot of ergonomic work to become street-legal, as it is difficult to get in and out, and requires such gymnastics as crossing your legs to reach the brake pedal if you're anywhere near average height.

But these adjustments feel like simple window dressing compared to the feats of engineering already achieved by this team to build this record breaking car. Smith identifies this as a big challenge, but making eVe Australia's first road-legal solar-powered car would be a record that can never be taken away from the team.

Tomorrow’s Fastest Cars Could Be Covered in Morphable Skins

As reported by Wired: Wrinkles aren't usually an aspect of the future that gets people excited. But fast cars are. And someday we might have cars that can accelerate more quickly, and efficiently, by morphing their surface texture through the mechanics of wrinkling.

Speed-enhancing body wrinkles on your Tesla are still years away, but researchers at MIT have created what could be the first step: a ball with morphable surface texture. They were able to get their creation, which they call a smorph (short for smart morphable surface), to wrinkle into a dimpled pattern similar to a golf ball’s, with similar aerodynamic properties.

Smorphs are sort of like raisins. As the soft inside of a grape dries out, the stiffer skin can’t shrink with it. Instead, it develops wrinkles to conform around the reduced volume. Smorphs don’t dry out (they also make terrible snacks), but the volume of  a smorph can be similarly reduced by sucking air out of its hollow core. That core is surrounded by different polymers: a thick, squishy layer covered by relatively stiff outer skin. As the core shrinks, the squishy layer is soft enough to contract smoothly, but the skin is forced to wrinkle.

The trick is controlling exactly how a smorph wrinkles. MIT mechanical engineer Pedro Reis, the material’s lead inventor, studies how wrinkling and other types of structural failures can be made useful. He says the first step toward controlling the wrinkling of a smorph is making the squishy base layer thick enough that the sphere doesn’t crumple like a ping pong ball. From there, they can tune the pattern of the wrinkles by changing the thickness of the outer skin. Dimples form when the skin is one-tenth to one-hundredth of the sphere’s radius.

The dimpling isn’t a pre-designed pattern. It’s the natural way the Smorphs’ skin collapses when air in the center is sucked out.  Pedro Reis/MIT

Because the smorph dimples look so much like those on a golf ball’s surface, the researchers were inspired to test their creation in a wind tunnel. It’s well established that a golf ball’s dimples help it fly further. Air passing over the dimples creates a bunch of tiny vortices. Rather than slowing the ball down, these vortices create a thin, turbulent sheath that the surrounding air can’t cling to. The result is lower drag. 

Sure enough, when the researchers tested the smorph in a wind tunnel, they found that it was about twice as aerodynamically efficient when dimpled.

But the sheath of vortices only forms at relatively low speeds. If a golf ball were to fly fast enough, it would be better off with a smooth skin. This is where smorphs could offer a huge advantage.

“What our system lets you do is tune the drag between the two extremes,” Reis said. Because of their size, golf balls rarely reach a speed when the dimples are less efficient. But something bigger, like a car, could be more fuel efficient with a few strategically placed morphable surfaces that would be dimpled at slower speeds and smooth when the car speeds up.

Earlier this year, Reis won an NSF grant to keep developing smorphs, which he hopes to someday scale up to use on cars, aircraft, and even buildings. He’s optimistic, but says this is probably a long way off. One problem is that hexagonal dimples are unstable on flat surfaces. So far smorphs have only been used on a round, ball shape, but Reis and his co-authors believe they can figure out how to reproduce the pattern on slightly curved surfaces. Creating the same aerodynamic dimpling on a car’s complex curves will be even more challenging.

Suit Against Walmart in Tracy Morgan Crash Claims Carrier Violated HOS Rules

As reported by  Overdrive Online: Four survivors of the June 7 New Jersey Turnpike truck crash that left comedian James McNair dead and actor Tracy Morgan in critical condition are suing Walmart Transportation, claiming the private carrier was “careless and negligent” in its oversight of drivers and equipment, which led to the crash, the lawsuit alleges.

Tracy Morgan is among those suing. The suit was filed July 10.

NTSB: Truck operator in Tracy Morgan crash was within
hours limits, but speeding.

The Walmart driver, Kevin Roper, reportedly was awake for 24 hours prior to the crash, though the National Transportation Safety Board in its preliminary crash report said he was within his hours-of-service limits, both in on-duty time and in driving time.

However, Roper had to commute in his personal vehicle 700 miles to the Walmart terminal in Delaware where he worked, the lawsuit notes and news outlets have reported. During the commute he was not on-duty. 

The lawsuit stems from Roper’s odd commute, alleging “Wal-Mart knew, or should have known, Mr. Roper was awake for more than 24 consecutive hours immediately before the subject accident.” The lawsuit alleges Roper fell asleep at the wheel, ultimately leading to the crash.

Walmart issued a statement responding to the lawsuit, saying it is cooperating with the investigation and is “committed to doing the right thing.”

“This has been a terrible tragedy,” the statement says. “We wish Mr. Morgan, Mr. Fuqua Jr., and Mr. Millea full recoveries. Our thoughts continue to go out to them, their families and friends, as well as to the families and friends of everyone involved, including Mr. McNair who lost his life. We are deeply sorry that one of our trucks was involved. As we've said, we’re cooperating fully in the ongoing investigation. We know it will take some time to resolve all of the remaining issues as a result of the accident, but we’re committed to doing the right thing for all involved.”

The suit also accuses Walmart of intentionally and regularly violating federal hours-of-service (HOS) limits for drivers, saying it “condoned this practice of its drivers routinely violating” federal rules.

The private fleet also failed to factor in driver commutes into its scheduling, the suit claims.

Driver pleads not guilty in crash charges, HOS argument brims.

“Wal-Mart had a custom and practice of recklessly and intentionally allowing its drivers to drive for prolonged and unreasonable periods of time, making them exceedingly vulnerable to suffer from fatigue,” the suit reads, adding that it also “recklessly and intentionally failed to take proper measures to combat” driver fatigue.

Lastly, the suit alleges the Walmart truck involved in the crash was equipped with an autonomous braking system, which did not deploy prior to the accident. Walmart “knew or should have known” the autonomous braking system was “compromised,” the lawsuit claims.

In addition to Morgan, comedian Ardley Fuqua Jr. is a plaintiff, along with Morgan’s personal assistant, Jeffrey Millea, and Millea’s wife Krista. The group — along with two others, including James McNair — were traveling in a Mercedes Sprinter van from one comedy show, where they were performing, to another.

They’re seeking actual, compensatory and statutory damages, punitive damages and attorney’s fees and court costs.

Roper has been charged with counts of vehicular homicide and assault by auto, to which he has pleaded not guilty. NTSB’s report concluded that Roper was traveling more than 20 mph over the posted 45 mph speed limit.

The crash also made the debate over the 2013 hours-of-service rule a national issue, causing lawmakers and celebrities to condemn Senate action in June to rollback certain HOS regs.

The Senate’s annual Department of Transportation funding bill includes an amendment that would suspend two of the 2013 rule’s 34-hour restart provisions, including the once-per-week limit and the requirement that the restart include two 1 a.m. to 5 a.m. periods.

New Jersey Senator Cory Booker proposed an amendment to strip that amendment from the bill. However, the bill has been idle in the Senate for nearly a month now.