Tech Talk

I’ve got you under my skin.
I’ve got you deep in the heart of me,
So deep in my heart, you’re nearly a part of me.

Clinical laboratory tests are like snapshot photos: you draw some blood, send it to the lab, and (eventually) get an impression of body’s metabolic condition at the moment the needle pierced the skin. How much cooler it would be if there were something like cell-phone video—a continuous succession of data on the fluctuations of key biological parameters, covering not just a moment or an hour, but weeks and even months. Such a device could transform both medical research and the clinical monitoring of chronic conditions like diabetes.

An interdisciplinary team at the Ecole Polytechnique Federale de Lausanne (EPFL) is one of the groups pursuing implantable wireless biosensors. At the Design, Automation and Test in Europe (DATE) conference in Grenoble, France, researchers Giovanni De Micheli, Sandro Carrara, and co-workers reported progress on their i-IronIC biosensor system. The device consists of an implanted miniature laboratory built into a tiny box just 2.2 by 2.2 by 15 millimeters and a skin patch that provides power, controls, and data relays between the patch and a Bluetooth-enabled cell phone.

Into the 0.07 milliliter implant package, the researchers have packed five customizable biomolecule detectors, along with monitors for pH and temperature. The pH sensor is based on iridium oxide; the thermometer is platinum. Each of the  biomolecule sensors is a three-electrode detector whose working electrodes are coated with a special layer comprising chitosan (an antibacterial long-chain sugar often used in implantable devices), multiwall carbon nanotubes, and an enzyme that catalyzes the molecule of interest.  The chemical reactions produce current flows within the detector. They are interpreted by a built-in microprocessor (the device is capable of both voltammetric and amperometric analysis) and then transmitted to the power-and-communications patch on the patient’s skin.

Current i-IronIC studies focus on glucose, lactate, glutamate, and adenosine triphosphate (ATP)—all components of the body’s energy production and consumption processes—but the developers say they can produce electrodes that will report for up to a month and a half on a wide range of metabolites.

For the last month, Bitcoin has been growing fat on the vine, like an over-watered grape in July. But the fun is now over. The fruit has burst and all of the juices are running out.

A massive sell-off began last night at around 4:00pm, yanking the exchange rate down from a high of US $260 to nearly $100 and reducing the total value of the bitcoins in circulation by more than one billion dollars. Mt. Gox, the largest of the Bitcoin exchanges, announced this morning that it will suspend trading until tomorrow. But the decline has continued on smaller exchanges, which have remained open. As of writing this, the price had already sank to $50 on the BitStamp exchange, setting Bitcoin back to where it was a month ago. If that rate holds, a whopping $2 billion will have evaporated in less than 24 hours.

The timing of the sell-off coincides with a spate of technical failures over at Mt. Gox. Yesterday afternoon, as the exchange rate rocketed up to $260, some traders began to notice a lag time on their orders. As the day went on, it got worse until many were complaining that their trades were taking two hours to go through. Meanwhile, the price began to drop rapidly and by the time orders were being filled, they no longer reflected the exchange rate.

Mt. Gox had a similar problem last week. After customers complained about not having access to their accounts, the company revealed that it had been the target of denial of service attacks. The exchange rate grumbled in response but ultimately continued its vertical surge. This time, however, Mt. Gox is not blaming an outside attack. According to an announcement on its blog:

"First of all we would like to assure you but no we were not last night victim of DDoS but instead victim of our own success."

Mt. Gox claims that it has been signing up 20 000 new accounts every day and had seen the number of trades triple in the 24 hours prior to suspending trading. It said it was this sudden increase in the trade volume that froze the trade engine.

As a result, it decided last night to suspend trading altogether until 12 April and will, it said, use this time to upgrade their system and catch up on the demand. Traders will be able to cancel their orders during this time and when trading resumes there will be no fees for 48 hours.

But this could be the beginning of a revolt against Mt. Gox, as Bitcoin holders take to online forums to vent their anger. They are especially dismayed that Mt. Gox continued to accept trades while fending off attacks by hackers. And many are calling for its customers to leave Mt. Gox and redistribute their money into smaller, competing exchanges.

Right now, Mt. Gox claims to handle 80 percent of all the trades in Bitcoin. Some kind of redistribution would seem fitting for a currency that was created to oppose centralized monetary controls.

Or, the answer may arrive by way of innovation. Bitcoin itself decentralized the concept of money. Perhaps the same could be done for an exchange. Mike Hearn introduced the idea of a peer-to-peer currency exchange at the last Bitcoin conference in London, and the people working on Ripple have come up with something similar.

One thing is clear from this latest debacle. In order to have a stable currency, the currency exchanges themselves must be stable and secure.

Image: Bitcoin Charts

Every ecosystem breeds parasites and predators. The Open Access Publishing movement, a high-minded effort to break high-priced journals’ copyrighted death grip on scientific information, is no exception.

The ranks of Open Access are growing. These journals –many of them good and a few excellent—make peer-reviewed papers available to all comers without subscription fees. Usually, they charge the authors to publish, rather than charge the readers to read.

Below the idealistic surface, though, lurks a new breed of opportunists out for profit alone. More and more, scientific researchers are being duped into submitting papers to journals that boast comfortingly austere names and noble Open Access mission statements, but exist mainly to charge authors exorbitant fees to publish articles that may never be cited or even read, as recent investigations—by Declan Butler in Nature and then Gina Kolata in the New York Times—have revealed.

Open Access is an appealing proposition. Scientific information should be available to whoever needs it, without expensive paper subscriptions, site-licenses, or Web pay walls. The doctrine fits perfectly with the Internet’s “information wants to be free” ethos. And it seems like simple justice, in a world where public money pays for so much research. Indeed, more and more government funding agencies are demanding that research underwritten by taxpayers must be published openly and without restriction.

In overturning the old business model, though, Open Access created new opportunities for profiteering. To be sure, no one can claim that the traditional model was without abuses: proliferating legions of unnecessary journals with high subscription rates, infinitesimal circulations, and even smaller citation rates; and unscrupulous sponsored journals that slapped a tinsel seal of approval on papers that are really bought and paid for. And it’s worth noting that both Nature (where I was once U.S. editorial director) and the Times have dogs in this fight. Nature has been the international flagship of traditional scientific publishing since 1869. And the Times has been struggling to hold its own as a generator of original content in the age of  aggregation.  Both are supported by subscription fees and advertising. Neither is thrilled about Open Access.

Taser has gone beyond stun guns by betting its future on a head-mounted camera worn by police officers. The 3.2-inch camera aims to help police departments reduce the use of force and defend against lawsuits regarding police brutality—as long as they can figure out camera-use guidelines that maximize oversight of police behavior while minimizing police surveillance.

The new Axon Flex camera can clip onto a police officer's baseball cap or sunglasses and operate continuously for up to 12 hours, according to a feature article by The Verge that coincided with the camera's official debut on 5 April. But the camera only begins active audio-and-video recording when the officer presses a small button on the control unit. The captured footage includes 30 seconds of video (without audio) from before the button is pressed. 

Any recordings get uploaded to a website called Evidence.com when the officer plugs his Axon Flex control unit into a docking base. That allows officers and administrators (high-ranking officers) to label and edit video files, even as the system records the user's login information and tracks the time he or she worked on the file. Video recordings stay in the system for only 180 days before being erased, unless they are used as evidence.

Such cameras could pay off for police departments by providing huge cost savings when defending against lawsuits, said Scott Greenwood, a national civil rights lawyer based in Cincinnati, Ohio, in an interview with The Verge. On the other hand, Greenwood envisioned a new world where judges would only consider police officer testimony when the officers could provide video to back up charges against defendants.

The new Axon Flex camera helped the Rialto Police Department in California dramatically reduce the number of "use-of-force incidents" and complaints during a yearlong trial that began on 13 Feb. 2012, according to a new study touted by Taser International. Chief Tony Farrar of the Rialto Police Department conducted the study as part of his graduate degree thesis at Cambridge University in the UK.

But turning every police officer into a walking version of the show "Cops" also comes with privacy risks and possible abuses of power. Greenwood emphasized that police departments must set up guidelines that let officers know when they should or should not use the cameras. For instance, police officers might be instructed to always turn the camera on when responding to certain scenarios, but would otherwise leave the cameras off to avoid extending police surveillance to ordinary citizens in normal circumstances.

Taser's venture into the new world of policing comes at a time when its fortunes have continued to rise and fall based on the controversy surrounding its flagship stun gun product. The stun gun's huge success in lowering the number of lethal confrontations between police officers and suspects has been marred by safety concerns surrounding a string of deaths and alleged cases of police brutality over the past years. If Taser's Axon Flex camera can encourage better policing and discourage abuse of both non-lethal and lethal force, everyone might end up feeling safer in the end.

Images: Taser International

This week Japan's cabinet approved a proposal that would reduce the power of the nation's "Big 10" utilities, which until now have had a firm hold on both electricity generation and transmission. By forcing the utilities to split their generation and transmission services into separate companies, the Japanese government hopes to increase competition, thus driving down energy prices and encouraging growth in the renewable sector.

The move is government's latest response to the Fukushima Daiichi nuclear disaster that destroyed four nuclear reactors, turned nearby municipalities into ghost towns, and terrorized the nation. Today, only two of the country's remaining 50 reactors are in service while the government tries to devise a new energy policy.

TEPCO, the utility that owns the Fukushima Daiichi plant, is the largest of the Big 10, which were set up as regional monopolies in the 1950s. The utilities invested heavily in nuclear power, and provided Japan with extremely reliable electricity. However, they also charged high rates and saw little reason to pursue innovations like renewable power. In the 1990s, the government began taking tentative steps to deregulate the industry, allowing independent companies to produce power and sell it to industrial and business customers.

We've profiled one of those independent businesses, Ennet Corp. Its CEO, Hiroaki Ikebe, explained that his company had no option but to use the Big 10's transmission networks, and that the utilities set very unfavorable terms for that use. The government's proposed overhaul of the energy sector will likely result in better terms for Ennet Corp and its like, as well as for renewable energy startups. The proposal would also open up new business opportunities for independent power producers, as it would allow them to sell power to residential customers as well. 

The government proposal must still be approved by parliament. It would be implemented over five years beginning in 2015, although it includes some wiggle room that worries those in favor of deregulation (for example, a clause says that the government must guarantee the stability of the energy sector before allowing the changes to begin). 

It's not yet clear how Japan's energy sector as a whole will be reshaped in response to the nuclear disaster. The prior administration advocated a full phase-out of nuclear power, but the current prime minister, Shinzo Abe, has endorsed reopening nuclear plants if their safety can be assured. While there's a great deal of enthusiasm about renewable power sources like solar, wind, and geothermal, such sources currently provide less than 3 percent of Japan's electricity generation. It's uncertain how much and how quickly that figure can be increased. 

Images: OiMax, Wikimedia

On Tuesday, after weeks of buzz in the neuroscience community, President Obama announced the BRAIN Initiative to map activity and connections within the brain. Obama's 2014 budget proposal will include $100 million to jumpstart this "big science" initiative, which builds on researchers' interest in understanding the neural circuits that are activated when we perceive, think, and act.

Though the U.S. government is already funding a similar $40-million venture called the Human Connectome Project, even the HCP scientists say the new program can fill gaps in current research. 

In his announcement, Obama compared the neuroscience initiative to the Human Genome Project that finished sequencing the entire human genome a decade ago this month. However, unless there are stunning and unanticipated breakthroughs in brain imaging over the next few years, the BRAIN Initiative won't result in a comprehensive map of the roughly 86 billion neurons in the human brain and the trillions of connections between them. In fact, its results may primarily illuminate the brains of fruit flies and zebrafish. 

BRAIN, the acronym, stands for Brain Research through Advancing Innovative Neurotechnologies; the name is fitting, say researchers, because the effort's real focus may be on developing new imaging tools that let scientists look at the brain in new ways. 

"The Human Connectome Project produces images at one resolution, using real-world technologies that exist today," explains Daniel Marcus, an investigator with one branch of the HCP who also heads the Neuroinformatics Research Group at the Washington University School of Medicine in St. Louis. "What Obama was talking about was, 'Let’s invent the next level of tools that enable us to look at the brain with a much higher level of resolution.'"

between individual neurons. "When we see something light up, it’s representing tens of thousands of cells," says Marcus. "There are also already existing technologies that can look at individual cells, or even dozens of cells, but there’s this massive range in between that we don't have the tools to look at." The BRAIN Initiative could build imaging tools that provide a certain Goldilocks-like resolution—neither too close nor too far. 

Images: J. Lichtman for the Center for Brain Science at Harvard University; David Van Essen for the WU-Minn HCP Consortium

It’s easy to infiltrate a semiconductor chip supply chain with counterfeits. The path from the original manufacturer to the final use is notoriously weak, especially for older chip models, which are often needed for military applications. There are different types of counterfeits: they can be falsely labeled, used, broken, actual fakes, or, as we are told this week, hacked to a specific purpose by the mob.

In a blog post Tuesday, two executives from IOActive, a computer and information security company, posited that the mob could easily enter the realm of chip counterfeiting and sell insidiously hacked chips with devastating results.

It’s not a new concern, but IOActive gives it a new twist with the gangster angle. They’re not wrong about the threat, but the company’s blog post smells a little like fear mongering.

Indoor navigation—the use of sensors and various local radio signals to help smart phones figure out where they are inside a mall, hotel, museum, or other large building—started getting very interesting last year, when a number of consumer electronics and communications companies joined forces to start working on an indoor navigation standard. Apple was not part of that group; in the fall the company launched its own general navigation software that didn’t include an indoor component. Probably a good thing, given all the other bugs in that software that Apple had to deal with.

At the time, some analysts suggested that Apple might be shopping around for a little indoor navigation startup to acquire. This week, Apple apparently found what it was looking for, acquiring Palo Alto’s WiFiSLAM, an alumni of Stanford’s StartX incubator, for $20 million. WiFiSLAM, started by a group of recent Stanford graduates and Google alumni, is just two years old; it uses existing Wi-Fi signals in a building and supplements those with what it calls trajectories. Trajectories are paths that it calculates from the existing sensors on phones as a user walks around a building, particularly, accelerometers, gyroscopes, and magnetometers. It saves these paths anonymously, and combines them, using pattern recognition techniques, to create maps of buildings. WiFiSLAM’s technology makes app users mapmakers, with their contributions used to make maps more accurate.

Meanwhile, Samsung is including pressure sensors in its Galaxy S4 form, according to business analytics company IHS. This means that phones will be able to tell how high they are in a building, not just in which direction a user is walking. Samsung is a bit ahead of the curve, IHS projects that Apple will add pressure sensors to its phones sometime next year.

Apple, it turns out, wasn’t the only established company to go shopping in Palo Alto in the past week or two. On 15 March Palo Alto startup Orchestra, another two-year-old company, was picked up by Dropbox for a rumored $100 million in cash and stock. Orchestra makes Mailbox, an app that simplifies email management on smart phones. And on 22 March, Trip Advisor announced that it had acquired Tiny Post, a company that makes an app used to add text to photos. Tiny Post’s app went viral when users began creating clever posters and sharing them on Facebook. No word yet on how much Trip Advisor spent for Tiny Post, also about two years old.

Video below: WiFiSLAM cofounder Joseph Huang explains the technology.

Spring is in the air. Here in Silicon Valley, it seems like just about everything is in blossom—daffodils, wildflowers, trees, and, it turns out, technologies. Last week a number of technologies that were, at best, tiny buds a few months ago have started to flower.

Qbotix. I first wrote about robotics company Qbotix last fall, intrigued by its approach to positioning solar panels to make the most efficient use of the sun. Instead of attaching each panel to a complicated motorized tracking assembly, Qbotix has built a robot, the SolBot, that runs on a track through a field of solar panels on simple stands; the robot figures out the best angle for each panel and turns it appropriately. Until now, the only SolBot in action was at Qbotix’ Menlo Park, Calif., headquarters, but last week the first commercial project went live—a 48 kw power plant at the Alameda County Santa Rita Jail in Dublin, Calif. The facility is using single Qbotix SolBot to position 32 panels, which are expected to generate approximately 120,000 kw hours of electricity per year.

Nanocoatings. At the past two International Consumer Electronics Shows (CES), startup companies wowed attendees by demonstrating nanocoatings that waterproofed personal electronics, invisibly making smartphones and pad computers impervious to at least a short dunk in a swimming pool. Last week I heard from a company called Semblant, that aims to take this kind of waterproofing technology into the industrial world, sealing electronics boards, solar panels, and entire cars (something all those folks who faced major damage to their cars’ electrical systems during Hurricane Sandy sure would have liked to have.)

4K Television. Remember those 4K TVs that several consumer electronics manufacturers were talking up at the CES this past January? The good news is that they really are going to be available this year, at least Samsung’s is, and you can preorder yours today. The bad news? The price. Back in January when I talked to manufacturers about possible pricing, no official information was available, but manufacturers were tossing around the $20,000 figure. Samsung this week announced pricing for its first 4K TV, an 85-inch model--$40,000.

Three years ago, an international team of scientists, led by researchers at the Max Planck Institute for Evolutionary Anthropology, in Leipzig, Germany published the first draft of the Neanderthal genome. Now the German group says they have computed a much higher quality genome.

The first draft was decoded using DNA fragments collected from three different bone pieces. The researchers have generated the new version from one toe bone, so it represents the genome of a single Neanderthal individual. They plan to publish a scientific paper later this year, but have already made the entire sequence freely available online for other scientists.

Computing the DNA blueprint of an extinct species is no easy task. Sophisticated DNA sequencing and computing techniques helped the team put together the first draft of the roughly 3.2-billion base-pair long genome (about the size of a modern human genome).

One challenge is that DNA fragments from fossil bones are typically only about 50 bases long; once these fragments are sequenced, assembly algorithms sort through the short sequences and string them together into longer and longer sections. During sequencing, though, some base positions get sequenced multiple times and others are missed completely. In the 2010 draft version, each position was determined once on average. New sequencing techniques the group has developed over the past two years have allowed them to sequence every position in the genome 50 times on average.

“Seeing each position that often dramatically reduces the chance that we make an error in the sequence,” says Janet Kelso, a bioinformatics researcher at the Max Planck Institute. “This 50-fold coverage Neanderthal genome is as good as, or better than the genomes that have been sequenced for many present-day humans.”

Here’s the caveat: when genomes are sequenced with next-generation sequencing technologies, some regions, typically those composed of highly repetitive sequences, simply cannot be confidently reconstructed, says Kelso. So these regions are generally not included in the final sequence.  

That’s why this ARS Technica article boldly, and rightly, says that the Neanderthal genome is not complete even though it’s about as good as we can probably get with prehistoric genomes.

But as Kelso points out, the problem exists for all genomes, be they old or new. “In this sense, I would argue that there is no complete human genome—modern or ancient!”

Photo: Nikola Solic/Reuters