Where Cellular Networks Don’t Exist, People Are Building Their Own

As reported by Wired: Inside the cloud that is perpetually draped over the small town of San Juan Yaee, Oaxaca, Raúl Hernández Santiago crouches down on the roof of the town hall and starts drilling. Men wearing rain gear of various impermeabilities cluster above him, holding a 4-meter-tall tower in place. Braided wires trail from four small circles welded near its midpoint; eventually those will be bolted or tied down in order to hold the tower steady during the frequent storms that roll through this part of Mexico’s Sierra Juárez mountains.

They don’t want it falling over every time it rains. Ninety thousand of the town’s pesos—a bit over $6,000—are invested in the equipment lashed to the top of the tower, in a town where many residents get by on subsistence agriculture.

The tower—which Hernández, Yaee’s blacksmith, welded
together out of scrap metal just a few hours earlier—is
the backbone of Yaee’s first cellular network.

 The 90,000 pesos come in the form of two antennas and an open-source base station from a Canadian company called NuRAN. Once Hernández and company get the tower installed and the network online, Yaee’s 500 citizens will, for the first time, be able to make cell phone calls from home, and for cheaper rates than almost anywhere else in Mexico.

Rhizomatica’s Peter Bloom helped make sure the tower could support a base station and an antenna.  Strategically ignored by Mexico’s major telecoms, Yaee is putting itself on the mobile communications grid with the help of a Oaxaca-based telecommunications non-profit called Rhizomatica. Its founder, Peter Bloom, is among the men currently getting soaked on the roof of town hall. It’s May of 2014, and this is the third of what he jokingly calls “artisanal cell phone installations” that he’s led in the Sierra Juárez in the past year and a half—the first of their kind in the world.

By the end of the year, he will have installed six more networks all over the state of Oaxaca, bringing the total to nine. Armed with an experimental concession from the Mexican government that grants Rhizomatica access to coveted cellular spectrum all over the country, Bloom is slowly but surely bringing coverage to towns that have been left out of the 21st century’s most important technological revolution.

Too small for profit

Of the world’s 7 billion or so cell phones, a few hundred of them are already in Yaee—they’re just not connected to a network. Kids use them as cameras and mp3 players, and Hernández, like many adults, bought his to use in Oaxaca City, a seven-hour bus ride away.

When he’s there, his cell phone can connect to plenty of base stations, which, in turn, link him to his choice of commercial network. But back in Yaee, there are no base stations and therefore no network. Every time Hernández wants to make a call in his hometown, he hikes for 20 minutes to the top of the highest hill around and hopes to catch some signal trickling in from a faraway base station, installed in a place deemed more profitable for telecoms than small towns like Yaee.

Raúl Hernández, el herrero de Yaee, construyó en su taller la torre hecha de chatarra.  Lizzie Wade

We’ve all heard plenty of uplifting stories of the democratizing potential of cell phones, how they’ve brought everything from voice calls to mobile banking to people who have never had access to landlines and laptops. Cell phones “have definitely proven the most ubiquitous piece of communication and digital hardware that people own on earth,” says Bloom. But on its own, “your cell phone doesn’t really know how to do anything,” he explains. All of the utility is in the network. And by and large, that network is provided—and, therefore, controlled—by a company that wants to make a profit.

That profit comes from subscribers, and if there aren’t enough of them in a particular region, cellular providers simply refuse to install their infrastructure there. Some countries get around that economic reality by legally requiring telecom companies to build networks in rural areas, no matter how many people end up paying for a contract. Mexico doesn’t have any such laws, meaning that Yaee, with its 500 residents, doesn’t stand a chance of attracting a commercial provider.

To make things worse, Mexico’s telecom industry is largely controlled by Telmex, a near-monopoly run by Carlos Slim. Ever since a supposed reform in the late 1980s transferred the country’s state telecom into Slim’s hands, Mexicans have paid first world rates for third world service—first for landlines, and now for cell service and internet access. And that’s when they live in a place with a network. Limited access and high prices meant that only 55 percent of Mexicans were using cell phones in 2011, according to the International Telecommunication Union.
Despite Mexico’s reputation for horrendous, Slim-driven telecom service and policy, it’s far from the only country that struggles with providing rural cell phone access. According to the GSM (for Global System for Mobile Communications, the standard technology behind a 2G network) Association, a consortium of commercial mobile providers from all over the world, 1.6 billion people in rural parts of developing countries don’t have access to mobile networks. That’s why Bloom and his collaborators at Rhizomatica say that if you really want to make the benefits of cell phones available to the people who need them most, it’s not enough to democratize the hardware by making the phones themselves super cheap. You have to democratize the infrastructure, the network itself. And that’s a lot harder to do.

Democratizing technology

From a hacker’s point of view, mobile communications came along at just the wrong time. The first commercial systems were deployed in 1991, right before the internet emerged from the academy and started making its way into people’s homes. By the time a strong open source community came into being, cellular networks were locked up behind walls upon walls of patents and proprietary equipment. Even today, “it’s very difficult to get your hands on the technology,” says Harald Welte, an open and free source software developer in Germany who works on mobile communications.

It wasn’t until around 2006 that old base stations started showing up on eBay, giving interested hackers like Welte a firsthand look inside the (albeit already outdated) technology that made 2G mobile networks possible. Out of straightforward intellectual curiosity, Welte snapped up a few and, four years later, he was able to make the first call on his reverse-engineered, open source network, dubbed Open BSC, referring to the base station controllers that coordinate traffic on a cell network.

Now, Rhizomatica is pushing Open BSC to its limits out in the real world. “We’re amongst the first people actually putting it into a live environment and using the shit out of it,” Bloom says. He first got interested in community cell networks when he was living in Nigeria and working with communities that were protesting the presence of oil companies in the Niger Delta. The activists there had cell phones, but thanks to the high cost of service—not to mention political forces that could monitor their communications or even shut down their network on a whim—it was difficult for them to share information with each other or with larger audiences. So Bloom decided to help them build what’s called a mobile mesh network, which connects cell phones directly to each other instead of routing calls through base stations or commercial networks. But the technology, which is mainly used in disaster relief situations like post-earthquake Haiti, proved to be too unreliable for everyday use. Sustained, real-world levels of traffic overloaded them, and the mesh networks frequently collapsed.

When a local man suffered several serious snakebites and needed antivenom right away, there was no signal.

A few years later, Bloom moved to Mexico to be with his now wife, who works with community radio stations in the Sierra Juárez. These villages wanted but couldn’t afford commercial cell service, and Bloom started thinking about a way to continue the project he had started in Nigeria. He decided to ditch the mesh network idea and went on the hunt for serious telecom technology, ultimately settling on Welte’s Open BSC as the strongest open source system. But since Bloom’s background isn’t in programming, he needed help installing the software on open source base stations he buys from NuRAN and another company called Fairwaves. He started enlisting the help of any experienced hacker who happened to pass through Oaxaca City on backpacking trips and the like. (One of them, an Italian, would eventually move to Mexico permanently to be part of the Rhizomatica team.) Today, Bloom spends much of his time personally driving the equipment out to villages like Yaee, getting soaked on as many roofs as he needs to in order to get the networks up and running.

The communities pay 120,000 pesos ($8,000 dollars) upfront for the equipment and installation, about one-sixth of what the commercial provider Movistar charges for a similar rural installation. Ninety thousand of the pesos go to buy the hardware, and the rest covers Rhizomatica’s time and expenses. Subscribers to the community network pay 30 pesos (about $2) per month for all local calls and texts, and the town keeps any profit left over after paying for electricity and maintenance. Thanks to a Mexican company called Protokol, which provides internet access all over rural Oaxaca, Rhizomatica can also hook up the town’s network to a voice-over-IP connection, which allows users to make very cheap long-distance calls to Mexico City and even the US, where many people have relatives. Once the network is installed, Yaee’s residents will be able to call the U.S. for 20 centavos (less than 2 pennies) per minute. A similar call from one of the town’s public landlines runs 15 pesos (about $1) per minute, a prohibitive cost for many residents.

Still, commercial networks have “20 years of headway” over the open source approach, Welte says, and Rhizomatica’s community networks can suffer from their distinct DIY feel. Bloom, Hernández, and the rest of the team must make sure to install Yaee’s tower above one of the town hall’s windows, so they can run an extension cord through it and plug the base station into a wall socket. That means whenever the power goes out in Yaee—which happens frequently, especially during the May-to-September rainy season—they lose the cell network, too. And until the town could raise enough money to move the entire installation to higher ground than the town hall’s roof (which happened in August 2014, three months later), there was no guarantee that Rhizomatica’s signal would be able to reach up the hillside to where Hernández and a good portion of Yaee’s residents live.

It’s this fundamental instability that causes the most frustration for users in towns that have had their community networks up and running for longer. In Yaviche, a similar sized town on the other side of the mountain from Yaee that installed its local network in September of 2013, Abi Martínez Ramos serves as the rural doctor and says that having any cell service at all has been a boon for emergency medicine. But when a local man suffered several serious snakebites and needed antivenom right away, “there was no signal,” he remembers. Someone had to physically find Martínez to administer treatment, just like in the old days.
Back in Yaee, it takes the team about 2 hours out in the rain to anchor the tower to the roof.

Raúl Hernández Santiago and Peter Bloom helped
install the first cell phone tower in the town
of San Juan Yaee, Oaxaca.

But it turns out the rain has caused a more serious problem than merely soaking everyone to the bone. Three cloudy days in a row have depleted the solar panels that power Protokol’s repeater antennas, knocking out Yaee’s internet access. With no internet, Bloom and his tech team can’t get the network online. They make plans to come back the following week to finish the job.

Despite its problems, an increasing number of communities in Oaxaca are eager to be part of Rhizomatica’s experiment, attracted by the low price and the promise of complete control over their networks. Keyla Mesulemeth Ramírez, who helps run the community network in Talea de Castro, a town of 2,000 that volunteered to be a Rhizomatica pilot project in the spring of 2013, fields one such inquiry in her office the day before the Yaee installation. Four men from a town called Yalahui have heard Bloom is in the area and they want to talk to him about installing a network in their village. They grow coffee, sugar cane, corn, and beans, and they’re tired of not being able to call home when they’re out in the fields. It’s annoying not to be able to call when you’ve forgotten your lunch, they say, but it’s downright dangerous when someone has an accident and needs help. Plus, everyone in Yalahui has family in Mexico City and the US, and they want to be able to call them without worrying about how much it costs. They’ve driven for 5 hours to talk to Bloom about a possible solution.

Mesulemeth is frank with the Yalahui men about the cost of the installation and Rhizomatica’s waiting list. But she’s sympathetic to their frustration of being left off the cellular grid. “Before, cell phone service was a luxury,” she tells them. “Now it’s a necessity.” She promises to put them in touch with Peter, who finds the group lingering over lunch a few hours later. He goes over the costs again, and says he’ll get out to Yalahui as soon as he can. Three months later, they’re making calls on a brand new network all their own.

Tesla to Hike Electric Car Output to 'a Few Million' by 2025

As reported by Reuters: Tesla Motors Inc (TSLA.O) plans to boost production of electric cars to "at least a few million a year" by 2025 from fewer than 40,000 last year, Chief Executive Officer Elon Musk said Tuesday.

Speaking at an industry conference in Detroit, Musk said Tesla may not be profitable until 2020. In addition, Tesla sales in China were unexpectedly weak in the fourth quarter. He blamed a misperception by city-dwelling Chinese consumers that they might have difficulty charging their electric cars.

"We’ll fix the China issue and be in pretty good shape probably in the middle of the year," he said.

Tesla shares fell 7 percent in after-hours trade to $190.22 from a close of $204.25 on the Nasdaq. During 2014, Tesla stock rose nearly 48 percent.

Musk, who last year said Tesla will begin phasing in "autopilot" features on its Model S sedan, predicted that the company will be first to market with a fully self-driving car, but likely not until after 2020. While Tesla may have a driverless car ready in five years, the vehicles may not receive regulatory approval for another two to three years after that, he said.

Musk also said the company's long-delayed Model X sport utility vehicle will be launched this summer, while the lower-priced, higher-volume Model 3 is on track for a 2017 introduction.

The Model 3 will be critical to Tesla's goal of reaching an annual sales level of 500,000 vehicles a year by 2020, a target which Musk also reaffirmed.
If Tesla hits its target of a few million vehicles by 2025, it would put the company on par with Fiat Chrysler Automobiles (FCAU.N), which sold 2 million vehicles last year in the United States.

Musk said Tesla likely would not achieve profitability using generally accepted accounting principles until the Model 3 ramps up to full production in 2020, although it may report non-GAAP profits before then as sales volume rises.

Musk told attendees at the Automotive News World Congress that "we could make money now if we weren't investing" in new technology and vehicles such as the Model 3 and expanded retail networks, Musk said.

On another topic, Musk said he was open to partnerships with retailers to sell Tesla vehicles, but not until after the company no longer has production bottlenecks.

"Before considering taking on franchised dealers, we also have to establish (more of) our own stores," he said. Musk said "we will consider" franchising "if we find the right partner."

He did not elaborate, but said Tesla "is not actively seeking any partnerships" with other manufacturers "because our focus is so heavily on improving our production" in Fremont.

Last year, Tesla delivered about 33,000 Model S sedans. Musk said the current wait for delivery is one to four months. Tesla already has presold every Model S that it plans to build in 2015, Musk said.

He said he did not see the Chevrolet Bolt, a low-priced electric car planned by General Motors Co (GM.N) for 2017, as a potential competitor to he Model 3.

"It's not going to affect us if someone builds a few hundred thousand vehicles," he said in reference to the Bolt, which GM expects to price to compete directly with the Model 3.

But "I'd be pleased to see other manufacturers make electric cars," he said.

Cyber Attacks Demonstrated on Autonomous Ground Vehicles

As reported by Net Security: Mission Secure Inc. (MSi), a cyber defense technology and solutions provider, and Perrone Robotics Inc. (PRI), a provider of robotic and autonomous ground vehicle solutions, announced a pilot project to demonstrate cyber attacks and protections targeted at ground vehicles. The University of Virginia Department of Systems and Information Engineering is sponsoring the pilot project.

As vehicles increasingly rely on automation, software and technology enhancements to run basic functionality, those systems serve as a potential safety risk when under cyber attack. Mission Secure uses a proprietary methodology developed by the University of Virginia with the Department of Defense for identifying the most consequential and easy to carry out cyber attacks on any system that a defense capability must address.

Proprietary software and hardware tools and best practices have been developed to support the methodology and to facilitate test and evaluation of the defense capability provided by MSi’s Secure Sentinel solution.

The pilot project will use MSI’s methodology and tools to develop and simulate realistic cyber attack scenarios against onboard control systems using a robust autonomous ground vehicle platform from PRI originally fielded for the 2007 DARPA Urban Challenge (“Tommy Jr.”), ideal for rapid prototyping and testing of this nature.

The goal of the pilot is to demonstrate how to identify vehicle safety threats malicious cyber attackers could use to easily compromise the vehicle’s key control systems and how these attacks could be detected and protected using MSi’s Secure Sentinel. Attack and protection scenarios are designed around several safety requirements of vehicles used on the road today.

“We are all excited about rapid innovation and automation for better efficiency and convenience as the world rapidly automates, but industry has not fully accounted for cyber security required for safety,” stated David Drescher, CEO of MSi. “The cyber assessment methodology, accompanying attack scenarios and protections being demonstrated in this ground vehicle pilot project follow results of a similar project involving autonomous air vehicles recently demonstrated by our team.”

PRI has been involved with a variety of pilot programs with the Insurance Institute for Highway Safety (IIHS) and Vehicle Research Center (VRC) in Ruckersville, VA. The team provides a unique understanding in test protocols and data collection.

PRI has a long history in development of fully autonomous ground vehicle solutions using their general-purpose mobile autonomous robotics software platform “MAX”. PRI has fielded automated vehicle projects for the DARPA Grand Challenges, various commercial applications, and is leading the development of an automated robotic vehicle test system for the Insurance Institute for Highway Safety (IIHS) in Ruckersville, Virginia.
Throughout the pilot project MSi’s Secure Sentinel with System-Aware Cybersecurity technology will be used to monitor, detect, inform and correct against various cyber attacks available to an adversary. System-Aware Cybersecurity and Secure Sentinels recently made news in December for the ability to monitor, detect, inform and correct against cyber attacks in an Unmanned Aerial Vehicle demonstration completed by the University of Virginia in collaboration with the Georgia Tech Research Institute.

“This type of collaboration with cyber defense technology providers will continue to be extremely important in validating the serious nature of potential cyber attacks on both autonomous and human operated vehicles,” stated Paul Perrone, President and CEO of PRI. “MSi’s Secure Sentinel technology along with PRI’s fully autonomous vehicle and automated vehicle testing technology continue to push the forefront of the next wave of autonomous vehicle technology and safety considerations for ground vehicles. Our technology has been automating vehicles since 2004 and we were happy to bring our expertise to these test scenarios.”

PRI’s “Tommy” line of vehicles were fielded for the 2005 DARPA Grand Challenge and the 2007 DARPA Urban Challenge. The company’s MAX software platform technology has been automating vehicles since 2004 including passenger cars, Polaris ATVs and drop in actuator kits for vehicle safety testing. The project has implications across both autonomous and manned vehicles and in a variety of industries including the automotive industry.

Electric-Car Pioneer Elon Musk Charges Head-On at Detroit

Tesla Motors CEO Elon Musk, shown last year, is pushing hard to sell 500,000 vehicles a year by 2020.

As reported by the Wall Street Journal: When Elon Musk , who loudly disdains the traditional auto industry, makes his first public appearance in Detroit in two years on Tuesday, it will be easy to see how much has changed since then.

Mr. Musk’s Tesla Motors Inc. is worth six times more in stock-market value. He is pushing hard to sell 500,000 vehicles a year by 2020, up from 90 a day in the third quarter. And giant auto makers are on a collision course with Tesla like never before, with General Motors Co. showing off a new electric car at the Detroit auto show on Monday.

Mr. Musk’s response? He says he doesn’t plan to change a thing, from his proclivity for F-bombs to double duty as chief executive of rocket makerSpace Exploration Technologies Corp., or SpaceX, to a hands-on obsession with the tiniest operational and car-design details at Tesla.

He calls himself a “nano-manager,” works about 100 hours a week and still runs the auto maker largely as he did before it sold the first Tesla Roadster in 2008.

“I have OCD on product-related issues,” he says with a laugh. “I always see what’s... wrong. Would you want that? When I see a car or a rocket or spacecraft, I only see what’s wrong. I never see what’s right. It’s not a recipe for happiness.”

In a speech Tuesday at an auto-show event, Mr. Musk is expected to criticize larger auto makers for not responding to Tesla even more aggressively. He denounces the rest of the industry as only halfheartedly trying to produce battery-powered cars for the masses, not just early adopters.

More than a decade after launching Tesla, the 43-year-old Mr. Musk has gained lots of believers in his brash, iconoclastic quest to run gas guzzlers off the road. Tesla is worth $26 billion in stock-market value, nearly half the size of GM or Ford Motor Co. GM sold more than 27,000 new vehicles a day in the third quarter—or about 90 every five minutes.


Panasonic Corp. will invest as much as $2 billion toward the $5 billion cost of Tesla’s new lithium-ion battery factory, under construction in Nevada. The state approved $1.3 billion in tax breaks and other incentives for Tesla, among the most ever in the U.S. for any company.

Futuris Automotive is getting a new factory the size of three football fields in Newark, Calif., ready to rev up production of seats and headliners for Tesla, which builds its Model S sedans nearby. Headliners are attached to the inside roof of cars.

“We decided to move into the Bay Area because we believe in what Tesla is trying to do,” says Sam Coughlin, Futuris’s top executive in the U.S. He plans to double the number of jobs at the factory this year, including engineers wooed away from Detroit.

Just 0.3% of the new vehicles registered in the U.S. since the start of 2012 are fully electric-powered, according to IHS Automotive, part of IHS Inc. Tesla had more than 20% of all electric-car sales in that period.

Tesla’s growing size is one reason why Audi AG , the premium-car brand owned by Germany’s Volkswagen AG , is working on a compact sport-utility vehicle designed to go 300 miles on an electric charge, according to Audi executives. Tesla plans to start selling its new Model X sport-utility vehicle in the third quarter of 2015.

GM’s new Chevrolet Bolt would be aimed directly at Tesla’s forthcoming Model 3 and capable of traveling 200 miles with one charge, according to people familiar with the strategy. A concept version of the Bolt will be unveiled at the industry’s biggest annual event, the North American International Auto Show, which starts Monday.

The Chevrolet Volt, a plug-in hybrid introduced by GM in 2010, has an electric range of just 38 miles, though next year’s model will go 50 miles.

Some Tesla investors are jittery, pushing the Palo Alto, Calif., company’s share price down more than 10% in the past three months, compared with gains by GM, Ford and major U.S. stock indexes during the same period.

Analysts at some securities firms are worried that the 40% slide in gasoline prices at the pump since April might hurt demand for electric cars. Tesla shares also are being hurt by concerns that a several-month delay in the Model X rollout will make it harder for Mr. Musk to hit the 2020 sales target.

But because of the unwavering intensity of Mr. Musk’s determination, investors have flocked to Tesla and shrug off the fact that the auto maker has never made an annual profit. They don’t care that the Model S usually sells for more than $100,000, or about double the company’s original projection.

In the first nine months of 2014, Tesla had revenue of $2.24 billion, up 60% from a year earlier. Its net loss grew to $186.4 million from $57.7 million.

“There are a lot of investors whose main thesis is based on Elon Musk and his vision for the future of vehicles,” says Colin Langan, an auto-industry analyst at UBS AG . He has a “neutral” rating on Tesla shares.

Cristiano Carlutti is more blunt. He ran Tesla’s operations in Europe before leaving in 2011 for startup auto makerQoros Auto Co., where he is a top executive. Tesla was “driven and is still driven by Elon Musk,” says Mr. Carlutti. “You take Elon out of the company, [and] the market cap would go down 80%.”

Mr. Musk says the auto maker has “certainly gotten a lot of attention, but we make very few cars, so the effect is small. The value that Tesla can provide is…an example of what is possible.”

He said when Tesla stock hit $280 last year that it was probably too high. According to the company’s latest proxy filing, Mr. Musk owns a 27% stake worth about $7.2 billion based on Friday’s closing price of $206.66.

With characteristic directness, Mr. Musk admits Tesla has no succession plan. The company also has suffered from growing pains because of his domineering presence.

Some high-level managers quit or were fired after clashing with the chief executive over Mr. Musk’s insistence on doing things his way, according to interviews with dozens of current and former Tesla executives.

During the launch of Model S production in 2012, Mr. Musk set up an office in the middle of the factory floor in Fremont, Calif., and took over when hiccups emerged. He told workers to buy USB cables at nearby Fry’s Electronics Inc. stores after a snarl delayed a shipment from China.

“If you are fighting a battle, it’s way better if you are at the front lines. A general behind the lines is going to lose,” he says.

Elon Musk, right, and Peter Thiel, shown in 2000, made a fortune
when PayPal was sold to eBay in 2002 for $1.4 billion. (AP)

Three weeks before the first Model S was delivered to a customer, Mr. Musk demanded that the sedans have larger rear tires because he felt they looked better, according to people involved in the process. The last-minute design switch also required complicated tweaks to the car’s anti-lock braking system and carried the risk of shortening the car’s driving range.

Tesla engineers protested that the electronics supplier might not honor its liability warranty if Mr. Musk went through with the bigger tires.

Mr. Musk didn’t back down. The engineers did. And the design change went through without a hitch, says someone who was in the room when the decision was made.

Current and former Tesla executives say few people are willing to disagree openly with Mr. Musk. Those who fall out of sync with him often don’t stay long. Turnover in the auto industry can be especially problematic because it takes several years to design and roll out new models.

Last month, the chief of Tesla’s operations in China, Veronica Wu, left the company after being there for only one year. Tesla hasn’t announced the reason for her sudden departure, and she didn’t respond to emails or other attempts to contact her.

Mr. Musk “is very, very demanding, and it’s not the right speed for a lot of people,” says Ricardo Reyes, who left Tesla in 2012 as its vice president of public relations and was rehired in November when his successor quit after about six months on the job.

Before that, Tesla went nearly two years without a director of the public-relations department.

The car maker’s CEO “pushes really hard, and it’s mission driven. It’s constant activity,” Mr. Reyes adds. “He used to say that he only wanted ‘special forces’ working for him. No normal people.”

Mr. Musk agrees that he can be exacting but denies that he fires employees without reason or cause. “I don’t like to fire people. I hate it,” he says. “The issue I’ve had is firing people too late, not too early.”

Media relations have been largely handled by Mr. Musk himself, who writes his own tweets and blog posts. Last Monday, he answered questions on social news forum Reddit Inc., telling readers about a favorite Winston Churchill quote about surmounting adversity: “If you’re going through hell, keep going.”

In the legal department, Tesla went through three general counsels from 2009 to 2012. It then had no general counsel for more than two years until Mr. Musk promoted his former divorce lawyer, Todd Maron.

Mr. Maron had been Tesla’s deputy general counsel. Mr. Musk says the turnover was an aberration, noting that SpaceX has had the same chief counsel for a decade.

Last spring, Tesla pushed out its sales chief in Europe shortly after the company started selling cars there. An executive overseeing Japan and Hong Kong was let go last summer. 

George Blankenship, the former Apple Inc. executive who came up with Tesla’s slick retail-store concept, left Tesla in 2013 after his responsibilities were scaled back. He didn’t respond to requests for comment.

Tesla’s chief engineer quit months before the Model S sedan was launched, while Mr. Musk fired the chief chassis engineer working on the Model X sport-utility vehicle, saying at the time that he asked for the resignation because “there wasn’t a good fit for him.” Tesla shares fell 17% after the auto maker announced the two departures but then rebounded.

Several people at Tesla say Mr. Musk has looked on and off for years for a chief operating officer who would be the auto maker’s second-highest executive. Outsiders approached by Mr. Musk include Andy Palmer, the former chief planning officer at Nissan Motor Co. who left the Japanese auto maker in September to become CEO of Aston Martin Lagonda Ltd., a person familiar with the matter says.

Tesla technology chief JB Straubel, chief designer Franz von Holzhausen, business-development chief Diarmuid O’Connell and finance chief Deepak Ahuja have been at Tesla for at least five years and are considered close to Mr. Musk.

But none of them is seen as likely to succeed him. Mr. Musk has indicated that he will stay in the top job for at least four or five more years.

Tesla often spurns job candidates who have “too much automotive experience,” says someone recently recruited by the company. During an interview, the person was told that Tesla considered it a major negative to have extensive GM experience on a résumé. Tesla didn’t offer the person a job.

People who have worked at Tesla say Mr. Musk’s ambition and vision are inspiring. But the hard-charging culture can be exhausting, they say.

Brett Foster, a former Tesla engineer who redesigned faulty retractable door handles on the Model S, recalls a profanity-laced email in mid-2013 in which Mr. Musk prodded managers to soften their communications with lower-level employees. “ ‘Don’t be an a—’ was the gist of it,” Mr. Foster says. “I actually kind of liked that he did it.”

Some managers tried to emulate Mr. Musk’s never-say-no style, but “without being as smart as Elon,” says Mr. Foster, who didn’t interact with the CEO regularly. “I struggled with the culture and am extremely happy I am no longer there.” He now works at Samsung Electronics Co.
Ryan Popple, Tesla’s finance chief from 2008 to 2010, says Mr. Musk’s relentless perfectionism has steered Tesla to where it is today. Mr. Popple, now the chief executive of electric bus makerProterra Inc., remembers when Mr. Musk demanded that Tesla figure out how to make aluminum body panels for the Model S, rather than hire an outside supplier to do the job.

The process is highly technical, tricky and expensive. “He was the only person in the room who thought it was a good idea,” says Mr. Popple. “There were a lot of people, including me, who thought: ‘I don’t know how this is going to work.’ ” Mr. Musk responded: “We have to own it.”

Tesla now makes all the aluminum body panels it needs in its own factory.

Why the US Navy Is Building a Five-Foot Robotic Tuna

As reported by Popular Mechanics: The U.S. Navy just completed its first round of salt water testing of GhostSwimmer, a five-foot-long prototype UUV (unmanned underwater vehicle) that looks and swims like a tuna. Some day soon, the Navy might deploy robotic fish to spy on enemy forces or protect friendly harbors and ships. 

Typical UUVs are torpedo-shaped and propeller-driven. But Project Silent NEMO, which began in 2013 at the Chief of Naval Operations Rapid Innovation Cell (CRIC), seeks to develop a drone sub that piggybacks nature. "We're fascinating by the biomimicry aspect of it—the fact that is a very non-conventional propulsion system," CRIC Director Cmdr. Benjamin Salazar says.

The tail of the GhostSwimmer, built by Boston Engineering, moves up and down and side to side as a form of propulsion. Thanks to its fish-like maneuvering, GhostSwimmer can navigate waters as shallow as 10 inches and make unusually sharp turns. "It can turn almost within the length of its own body," Salazar says.

But there's performance in the lab, and then there's performance in the field. This most recent test looked at how GhostSwimmer survived realistic conditions at sea. "Very few experiments survive first contact with salt water," he says. "The sea is a harsh mistress."

Capt. Jerome Lademan, project lead for NEMO, tells PM that the robot fish is still at Technical Readiness Level 6—that's engineer-speak for a prototype not yet ready for prime time. To get there, the team will need to improve the software of UUV's navigation system, which relies primarily on an inertial navigation unit when the GhostSwimmer is untethered.

The engineers also need to extend GhostSwimmer's battery life, both by finding better batteries and making its propulsion system more efficient. The goal, Lademan says, is to "make each flip of the tail more effective in different types of water with different levels of buoyancy." In the summer of 2015, GhostSwimmer is scheduled to undergo tests of its endurance and ability to avoid obstacles. Right now, Salazar says, the UUV can travel 10 miles before needing to recharge.

The Department of Homeland Security is currently developing a variant of GhostSwimmer called BIOSwimmer, which has a ducted propeller at the end of its tail. DHS wants to use this bot for hull inspection and harbor security. Salazar insists that CRIC isn't looking at any specific missions for the robo-fish, though Boston Engineering notes that GhostSwimmer could be used for intel and recon missions, shallow water mine countermeasures operations, explosive ordnance disposal, hull inspections as well as "covert sentry missions."

 

SpaceX Rocket Makes Crash Landing On A Drone Barge: "Close, but no cigar this time."

As reported by Gismodo: After a lot of publicity and one aborted launch attempt, SpaceX tried to land its Falcon 9 rocket on a drone barge this morning. 'Tried' very much being the key word.

Initially, the mission went to plan: the rocket launched OK (incidentally, the first successful American resupply launch since that other rocket went boom), the cargo part separated from the main rocket first stage, and is now on its way to the ISS, where it should arrive on Monday.

The first stage made it to the 'autonomous spaceport drone' (that's a robot barge to you and me) without problem, but then, according to SpaceX founder Elon Musk, "landed hard":

But as Musk points out, the mission wasn't a complete failure. Landing a rocket vertically on a moving ship isn't exactly easy — SpaceX has likened it to "trying to balance a rubber broomstick on your hand in the middle of a wind storm". The actual landing is comprised of a few different parts: First, a boostback burn gets the craft headed in the right direction; a second 'supersonic retro propulsion' slows the rocket to 250 meters per second; and a final landing burn brings it to about 2 meters per second.

From the sounds of things, the first two parts went well: the rocket certainly hit its target, and presumably didn't do so at several thousand miles per hour. Two outta three ain't bad.