jul 2013 / last mod oct 2016 / greg goebel

* 23 entries including: waste management series, manufacturing robotics, climate sensitivity downgraded, smart grid for Bornholm Denmark, eye tracking, bitmessage, NSA security extensions for Android, Figaro the clever cockatoo does tools, tracking your kids, US government does tweets, hauling oil by train, crowdsourcing patent validation, and virtual humans for crash testing.

banner of the month

[FRI 26 JUL 13] WASTE NOT (11)
[FRI 19 JUL 13] WASTE NOT (10)
[FRI 12 JUL 13] WASTE NOT (9)
[FRI 05 JUL 13] WASTE NOT (8)


* NEWS COMMENTARY FOR JULY 2013: The recent flap over US government data-mining via the NSA PRISM program, discussed here last month, has some suggestion of a public policy train wreck. Suddenly, the public has become agitated over what amounts to well-established practices, some of which can be argued as prudent; others which are questionable, but which citizens have blandly accepted in their online activities. It's all more smoke than fire, but the smoke is still a problem, and such fire as there is needs to be dealt with. As confused as things are right now, events have provoked a degree of public discussion that should hopefully sort things out, at least to a degree.

As reported by BBC WORLD Online, as part of this discussion Apple, Google and dozens of other technology companies have urged US authorities to let them divulge more details about security requests. The companies want to be able to report regular statistics about the nature and scope of what data is being asked for. Companies are currently allowed to release limited data regarding security requests and their nature -- but as things stand, those disclosures must be limited in scope, and in many cases require that the firms ask the courts for permission to make the information public. Campaign groups such as the Electronic Frontier Foundation and Human Rights Watch are also endorsing the action.

The group sent a letter to the White House, Congress, and the National Security Agency. General Keith Alexander, boss of the NSA, sounded a relaxed note in response: "We just want to make sure we do it right. We don't impact anything ongoing with the FBI. I think that's the reasonable approach." He had no objection to the request by the companies: "What they want is the rest of the world to know that we're not reading all of that email, so they want to give out the numbers. I think there's some logic in doing that."

Not so incidentally, on 24 July the US House of Representatives put the future of PRISM up to a ballot; PRISM survived on a 217:205 vote, after the Obama Administration threw in weight to support it. The close vote underlines that the political feud is far from over.

* An article from THE ECONOMIST Online asked the question: Is tax reform in the USA possible? That's not a simple matter, but Democratic Senator Max Baucus and Republican Congressman Dave Camp are operating on the assumption that it can be worked out.

They're up against discouraging recent history. In 2012, President Barack Obama and Republican Congressman John Boehner, the Speaker of the House of Representatives, suggested in the course of their negotiations that a deal was likely. It didn't happen. However, Baucus and Camp, who head the tax-writing committees in their respective chambers, want to try again.

They both have strong motives for doing so. Baucus will retire at the end of 2014; tax reform would be going out on a high note. He's held 30 hearings in three years and published ten papers detailing concepts for tax reform. Camp is close to timing out as well, his term as chairman of the House Ways and Means Committee ending next year. Camp took up the cause in 2010; since then he has published three draft proposals on tax reform, held 20 hearings and formed 11 working groups. The two men meet weekly, have a joint Twitter feed and often travel together. Both want to address the tax code's two big problems:

American corporate tax rates are the highest in the rich world: add state and local taxes to the 35% Federal rate, and they come to 39.2%. However, the tax rules are so complicated that the government sees little of that money. Even Obama, no foe of corporate taxes, thinks they should be lowered, but also rationalized.

Camp wants to produce a bill by the end of 2013. He proposes to cap the corporate and individual rate at 25%; exempt most foreign profits from American tax, with safeguards to stop firms from shifting intellectual-property income to tax havens; and in general simplify the tax code. Baucus is taking a less specific but bolder approach; he and the top Republican on his committee have suggested as a straw proposal that all tax expenditures be ended, with advocates then to justify which should be retained -- with the recognition that any tax expenditure retained will raise tax rates proportionally.

Unfortunately, for every tax expenditure there's a faction that benefits from it; the only way to get around that issue is try to play them all off against each other, which may be easier said than down. Another issue is that Baucus wants to raise more revenue, while Camp wants to reduce rates -- though given their good relationship, they should be able to come to an agreement.


The biggest problem is that Republican leadership is strongly against any increase in taxes, and is solidly opposed to the Democrats' desire to raise revenue. The White House is firmly committed to cutting tax expenditures without lowering rates, which Republicans entirely reject. An opportunity for reform may come in October, when the US Treasury again hits its legal borrowing limit, the "debt ceiling". In the best of all possible worlds, Republicans would agree to more revenue while Democrats would agree to holding down the growth of entitlement programs, such as Social Security and Medicare. Nobody's betting on that happening.

* In related news, a blog posting from THE ECONOMIST passed on the findings of David Fahrenthold and Lisa Rein, reporters at THE WASHINGTON POST, concerning the Federal budget sequestration that went into effect in March. The administration warned of dire consequences, the two reporters writing:


There would be one-hour waits at airport security. Four-hour waits at border crossings. Prison guards would be furloughed for 12 days. FBI agents, up to 14. At the Pentagon, the military health program would be unable to pay its bills for service members. The mayhem would extend even into the pantries of the neediest Americans: Around the country, 600,000 low-income women and children would be denied Federal food aid.


Then they added: "But none of those things happened." On examination, they found that of the administration's 48 dire prophecies, half didn't come to pass, the jury's still out on 13 -- and of the 11 that have come true, only a few seem to give cause for serious concern.

There's a certain lopsidedness in the Federal tax system, in that for the most part any one program has little effect on how much a taxpayer forks over, a few pennies at most, meaning most don't have too much reason to resist a specific program -- but each specific program has its clique of beneficiaries, who will fight to preserve it. The pressure for growth well outbalances the pressure to cut, and so the government just keeps growing.

The idea behind the sequester was to take a meat cleaver to the overall budget, cutting enough to make it of interest to the taxpayers and their representatives. In other words, as inelegant as the approach seemed, if we were to cut spending, was there any other practical way to do it? And, as it turned out, it wasn't as painful as it seemed up front because, as Fahrenthold and Rein explained:


In some cases, agencies dug into their budgets and found millions they could spare. In other cases, Congress passed a law that allocated new funds or shifted money around. In others, lawmakers signed off on an agency's proposal to "reprogram" its money. In the process, the "meat cleaver" of sequestration often became a scalpel. It spared crucial programs but cut second-tier priorities such as maintenance, information technology, employee travel and scientific conferences.


There have still been losers, with some small programs for the poor taking a hit, and the Pentagon being put under particular pressure. To be sure, the public always tends to prefer butter over guns -- but the military has its operational commitments, and it doesn't make much sense to cut the defense budget while not taking anything off their plate. Worse, military forces have to be prepared for future, and likely unexpected, contingencies. There is also the unfortunate fact that putting off maintenance of military facilities and gear simply defers the final bill to a later date, and may well inflate the bill in the end. Such inflation is absolutely evident in military procurement programs, notoriously prone to perversities, one being that stretching out such programs is guaranteed to pile up costs.

The sequester was a hamfisted approach to budget cutting, not the rational way to address the issue. However, given the difficulty of tackling the Federal budget, maybe the sequester is just one of the tools we end up having to use -- if at the very least to encourage better ways of doing things.



* LITTLE GREEN GRID: There's been considerable interest in "smart electrical power grids" or "green grids" in recent years, concepts being discussed here in 2009. As discussed by an article from IEEE SPECTRUM ("The Smartest, Greenest Grid?" by Jean Kumagi, 29 April 2013), the Danish island of Bornholm is at the leading edge of green grid technology.

Bornholm, with 41,000 residents, is a pretty place, with a thriving tourist industry, along with arts and crafts, dairy farming, and commercial fishing. The island traditionally obtained its power from a 60 kilovolt, 70 megawatt (MW) cable running along the bottom of the narrow channel between the island and Sweden, tapping off the Nordic power grid. When the cable was working, the islanders had the power they needed, but it had an unfortunate tendency to be cut by ship's anchors and the like every few years, leaving the islanders literally in the dark while time-consuming repairs were made.

Of necessity, the islanders have worked for "energy independence". At present, Bornhold has about 50 MW of domestic capacity, from a mix of conventional coal and diesel generators, three dozen wind turbines that dot the countryside, rooftop photovoltaics, a biogas plant, plus several wood chip and straw fired plants. Today, a blackout only lasts a few hours, the time it takes to get local generation systems online.

Bornholm smart grid, done the Dane way

Now the islanders are moving on to integrate their power systems, through an exercise named "EcoGrid EU". Funded to the tune of 21 million euros ($27 million USD), with assistance from the European Union, EcoGrid EU will represent a new generation of power grid. It is an important exercise for the Danes, since in 2012 the Danish Parliament approved a plan to have renewables supply 35% of the country's total energy needs -- not just electricity, but also heating and transportation -- by 2020, on the way to 100% by 2050. That's extremely ambitious, and there's no way to achieve such targets without a power grid that can handle renewable energy sources.

The difficulty with renewables relative to a power grid is that they're inconstant, meaning a power grid has to be able to respond quickly and intelligently to changes in supply and demand. Bornholm's wind turbines can, on a good day, provide up to 30 MW of power, better than half of the island's peak load of 55 MW. However, wind can abruptly die off, causing the electricity supply to die off with it, and in particular changing the frequency of the power grid from its designed 50 hertz. Out of phase power from multiple sources can wreak havoc on a power grid -- causing power delivered to any one location to surge or sag -- and so the grid has to be shut down if the frequency goes too far off nominal, resulting in a blackout. To be sure, Bornholm's conventional power sources can take up the slack, but they take time to get going.

Along with the simple challenges of supply, there are issues of economics; of the relative costs of different forms of power available to Bornholm, along with the variation in costs over time; and of the ability of Bornholm to either take power from the Nordic grid, or sell power back as available. That's where EcoGrid comes in.

EcoGrid personnel installed smart grid controllers in about 1,200 households and a hundred businesses. Since April 2012, the controllers have been receiving a continuous stream of data based on the five-minute price for electricity in the Nordic electricity market. The controllers communicate via wireless with designated appliances, with algorithms determining whether to turn each one on or off -- based on factors like the time of day; the weather; along with current, past, and future market prices.

At first, the project's organizers envisioned regulating a whole suite of household machines -- dishwashers, washing machines, refrigerators, TVs, lights. The problem is that though smart appliances have been around for years, they can't communicate well because they don't work on a common wireless standard. For now, the primary appliances are electric heating systems and heat pumps. In 700 households and most of the businesses in the system, the heating system is directly controlled using algorithms developed at IBM's research lab in Zurich -- working off models for each site, based on factors like electricity usage patterns and the size of the windows and walls.

Another 500 or so households are being treated as single electricity-consuming units; Siemens' Danish subsidiary is coordinating that part of the exercise. The remainder of the households in the EcoGrid are just getting smart meters, which provide them with fine-grained information about their electricity consumption and market prices but don't control their usage. All participants can observe their energy use at any time on the system website.

Again, EcoGrid is not entirely focused on energy efficiency as such; load balancing is the major driver behind the effort. Go to a local supermarket and the refrigerators packed with frozen foods seem perfectly normal, but they have a special controller unit that monitors line frequency. If the line frequency changes too drastically, the refrigerators shut off for a moment to allow the line frequency to normalize. The idea of using appliances to balance electrical grids is not new, but it's only been recently that appliances have become smart enough to do the job.

Participants are being told not to expect much in the way of savings, studies of energy efficiency efforts in the past suggesting that payback to consumers is generally not proportional to the effort. However, the citizens of Bornholm like the idea, Danes being environmentally conscious, and there's been no shortage of participants. Says one: "In the future, we won't have that much power. And my son is probably going to have kids as well. Where are they going to get all the power from?"



* MANUFACTURING ROBOTICS ENERGIZED (3): Facilities crammed with robots are an attractive vision for company owners; workers and labor unions aren't so enthusiastic, seeing robots as taking away jobs from humans. The reality is a bit mixed. Boeing is struggling to find enough hands to produce its new 787 jetliner; the major impact of the company's robots is to improve productivity and quality for those workers. Earthbound's robots are taking away jobs, however, with each machine replacing two to five human workers.

The Obama Administration is enthusiastic about robotics, saying that it offers a historic opportunity for the USA to stay competitive. Tom Kalil, deputy director of the White House Office of Science and Technology Policy, commented: "The only way we are going to maintain manufacturing in the US is if we have higher productivity."

US government officials and industry executives argue that even if factories are automated, they still are a valuable source of jobs -- with the quiet subtext being that the choice is not between robots and jobs, it's between robots and no jobs anyway. Businesses have no legal or moral obligation to hire anyone, if a robot is the cost-effective solution to a problem, they'll buy a robot. If a business doesn't adopt cost-effective robots, someplace else will, the business will not be able to meet the competition; like the scenario or not, there's nothing anyone can do about it.

There's a fear that if America's manufacturing base disappears, product engineering and design will as well. On the plus side, a US boom in robotics will produce jobs for skilled workers to create and operate the robot factories -- but there will be little place for unskilled or semiskilled labor in the factory of the future.

* The continued creeping automation of labor suggests a future where any routine job will be done by a machine. To be sure, humans would still have work in creative and managerial positions -- after all, if we had a machine that could decide what to do and not merely be assigned to do it, we'd have to give it citizenship rights -- but most of today's jobs would cease to exist.

One Noah Smith, in an essay from THE ATLANTIC ("The End Of Labor", 14 January 2013), speculated on the implications of this scenario. His first premise is that a highly automated society would be an extremely productive and prosperous one, so there would be no shortage of money to go around, it would be just a question of its distribution. It would present plenty of opportunities for the enterprising, since a single individual or a small group could establish an automated business at low cost. But what about the rest of the people? Smith had a modest suggestion:


Everyone is born with an endowment of labor; why not also an endowment of capital? What if, when each citizen turns 18, the government bought him or her a diversified portfolio of equity? Of course, some people would want to sell it immediately, cash out, and party, but this could be prevented with some fairly light paternalism, like temporary "lock-up" provisions.


This sounds like Karl Marx with a degree in robotics, and Smith took pains to qualify the speculative nature of his remarks, describing it as "an extreme measure, for an extreme hypothetical case." Robots may end up augmenting human labor instead of replacing it; technology may never be able to replace humans in jobs requiring some level of skill and mental acuity. Still, we are faced with the possibility of a future where the machines do all the dirty-hands work, where there's an ocean of wealth to go around, but not so many jobs as such. The world doesn't owe anyone a living? In the future, maybe it will.

* As a footnote to this series, an article from THE ECONOMIST ("Baxter Gets To Work", 29 September 2012) highlighted "Baxter" -- a new, flexible industrial robot from Rethink Robotics in Boston.

Baxter was developed under the direction of Rodney Brooks, once a Massachusetts Institute of Technology robotics guru, best known for creating ingenious insectlike robots, and later a co-founder of iRobot, maker of the Roomba vacuuming robot. His Baxter looks a bit like a theme-park prop, being a two-arm robot on a rotating pedestal, with a flat-panel display for a head, the display providing a cartoony face of two eyes and eyebrows.

Industrial robots tend to have to be isolated behind barriers so they won't smack people passing by. Baxter is designed to interact with people safely. The robot's arm joints are not driven directly by electric motors, instead being driven by motors through springs, giving the arms a degree of elasticity and, in conjunction with force sensors, the ability to sense if they've hit something. Baxter also has a sonar system in his head to check if anyone's near. His cartoony face is also essentially for safety, giving an expression of surprise if anything unexpected happens, and also looking towards his next target of movement so people have an idea of what he's going to do next.

Baxter the robot

Baxter is easily programmed by moving its arms and manipulators through the proper motions, such as picking up items off a conveyor belt and boxing them. The display provides status data to help in programming. Baxter has five cameras to allow him to adjust his motions as items show up in different positions, or if he drops something.

At $22,000 USD, Baxter is cheap for an industrial robot, the price being paid off in a few years. Brooks thinks that the low cost of the robot will persuade industries that have never used robots to give Baxter a try. Rethink Robotics will provide regular software upgrades to make Baxter smarter, and the firm is introducing a developer kit to allow others to see what they can do with the machine to. [TO BE CONTINUED]


[FRI 26 JUL 13] WASTE NOT (11)

* WASTE NOT (11): Most schemes now being envisioned for nuclear waste disposal focus on underground burial. Salt deposits excavated 650 meters (2,130 feet) beneath the desert near Carlsbad, New Mexico, already store canisters with enough radioactive waste from America's weapons programs to fill more than 28 Olympic swimming pools. Burying spent fuel is trickier; unlike weapons waste, power-station waste generates lots of heat which might, in the long term, destabilize geological formations.

No country has yet built a permanent burial site for spent fuel, but the technology for doing so appears to be available. To find out how a repository carved out of argillite, a sedimentary rock, would hold up, French researchers built an underground laboratory 490 meters (1,605 meters) down near the northeastern town of Bure. Heat and radioactivity testing show that a repository could be built in about ten years and then safely filled and monitored for at least another hundred before being sealed.

Any permanent repository would be built in a dry geological medium, but water could still seep in through cracks created by, say, an earthquake, or shifting land mass, allowing radioactive material to seep out. To prevent that, Czech engineers working with a German consultancy named DBE Technology have built a "hydraulic cage" to keep water out of a former limestone mine near the Czech town of Litomerice that holds radioactive waste in concrete-lined, zinc-coated drums. In 2005 they began lining the top, bottom and sides of some of the chambers with a five centimeter (two inch) thick layer of gravel, held in place with wire mesh sprayed with concrete. The gravel layer's function is simple, merely providing a path for water to seep around the chamber instead of through it.

Finland and Sweden have also built underground laboratories to test geologic formations and packaging technologies, with positive results. Repositories built at depths of 500 meters (1,640 feet) or more in appropriate rock would even survive the glaciers of a future ice age as they scour Europe's landscape.

Despite the technological optimism, not one permanent repository has been approved. Such a facility is bound to be expensive, but public opinion is a bigger factor. Officials in France and Sweden don't see a lot of opposition, and believe that that building permits will be granted within a few years. However, in the USA public hostility helped derail the Yucca Mountain facility after more than $9 billion USD had been spent on it. It appears that the hostility was mostly due to inept public relations. People living near the Carlsbad site for weapons-related waste were consulted and compensated, so they didn't raise a fuss; Sweden's SKB chose the site for its proposed repository after a long consultation and several referendums.

America's Blue Ribbon Commission concluded that a "new, consent-based approach" was needed. The commission also called for research into an alternative form of underground burial -- packing waste into holes drilled several kilometers deep. Research into "borehole disposal", as it is known, is now picking up. Waste canisters dropped into the holes could be designed to melt the surrounding rock, which would eventually solidify and form a tight seal. A workshop held last October at Sandia National Laboratories in Albuquerque, New Mexico, concluded that borehole disposal would be cheaper, more flexible and faster to implement than repository disposal. The only problem with selling the idea to the public is that it may not be a good idea to emphasize that the wastes are hot enough to melt rock.

* As noted earlier, some believe that burial is the wrong solution. Andrey Bukhovtsev of Rosatom, Russia's state nuclear giant, reflected a common Russian attitude when he commented: "Spent fuel, in general, is not waste."

That reflects the optimistic view that pollution is nothing more than a misplaced resource. Reprocessed waste yields usable uranium fuel, along with what has been called "an alchemist's dream" of other elements useful in industry and medicine. France and Japan are big in reprocessing. It is expensive, but reduces waste and provides a secure fuel supply.

The problem is that the "alchemist's dream" of isotopes obtained from reprocessing also includes plutonium that could be used to build weapons. As a result, US President Jimmy Carter halted reprocessing in America in 1977, with some other countries following suit. Scientists at the US Oak Ridge National Laboratory have been tinkering with a new reprocessing technique called UREX that does not produce pure plutonium. However, it's not ready for operational use at this time, and even the optimists admit some waste will still be left over. The consensus is that we need to start burying it sooner instead of later.

* As a footnote to this series, AAAS SCIENCE NOW discussed one of the problems with long-term storage of nuclear waste: making sure that people are informed of the specifics of the storage scheme thousands of years down the road. Building a data storage system that can predictably last that long is a challenge.

The French nuclear waste management ANDRA has developed a potential solution: a sapphire disk with data laid down in platinum. The prototype costs 25,000 Euros to make, but it is designed to survive for at least a million years. It's actually made from two thin disks of industrial sapphire, each about 20 centimeters (4 inches) across. On one side of each disk, text or images are etched in platinum, with the two disks then fused back to back. All that's needed to read it is a microscope; the double-sided disk can store about 80,000 pages.

There's one big problem: what languages to write the disks in. Human writing hasn't been around for even 10,000 years, and we have no idea what language will look like 10,000 years from now, much less a million years. ANDRA has brought together researchers from a wide range of fields -- archivists, archaeologists, anthropologists, linguists, and even artists -- to get suggestions. [END OF SERIES]



* GIMMICKS & GADGETS: As reported by POPULAR SCIENCE, an inventor named Francisco Aguilar -- in charge of a startup in Cambridge, Massachusetts -- is developing a device for emergency response workers named the "Bounce Imaging Explorer". It's a black ball, about the size of a baseball, with a shock-resistant shell and fitted with six cameras around its surface. To check out a potentially hazardous or obstructed room, an emergency worker just tosses the Bounce inside, with the gadget transmitting imagery over wireless to the emergency worker's smartphone or tablet. Each camera port is flanked by four LEDs to provide illumination. The Bounce also relays carbon monoxide, methane, and temperature levels, and has a microphone to allow the emergency worker to hear what's going on. Aguilar plans to price the Bounce between $500 and $1,000 USD.

* I got a tipoff from WIRED Online for a smartphone game, XMG Studio's GHOSTBUSTERS PARANORMAL BLAST. The idea is that it gives players a map of locations where spooks are lurking; the players find the haunted spot -- park or laundromat or whatever -- and then scan the area with their camera phone, the pesky ghost then being displayed as an animated overlay on the realtime camera image. The game then provides a proton blaster and a ghost trap to make the catch.

"WHO YA GONNA CALL?!" Well, no need to call any more, now anybody can be a Ghostbuster. Sounds like fun; it almost makes me wish I actually had a smartphone.

* In personal gimmick news, my house doesn't have air conditioning, so I've tended to find summers wearying. The temperatures in northeast Colorado are not so extreme and it rarely gets very humid, so I've long wanted a small portable evaporative cooler -- it would do the job effectively and at low cost. However, I've never found a small one that was designed to be maintained, and since they can't be cleaned, they quickly mildew up to a rank state.

This spring, I was checking around at the home goods store to see if they had any evaporative coolers that were more promising. I didn't find them, but I did notice the store was carrying portable air conditioners. They were news to me, I'm sure they've been around for a while, but I never noticed them before; they're units on wheels that can be rolled around the house. I've never wanted to buy a window air conditioning unit because the installations seem klunky, but the portable unit seemed promising; it was inexpensive, and when not in use, easily stored.

I went back home and did some poking around online, the general opinion of people who bought portable air conditioners being that they are noisy -- but it's a white noise, and it's not hard to get to sleep with it -- and not very efficient -- but I only wanted to cool my bedroom-office, where I spent the bulk of my time, so that didn't seem a worry. I promptly went back and bought an LG LP1013WNR 10,000 BTU air conditioner. LG, was that a Korean or Chinese firm? I looked it up, and it was the Korean firm originally known as Lucky Goldstar.

Setting up the air conditioner was straightforward. It had an extensible panel that could fit vertically in my window; the air conditioner had a kit of stick-on gaskets, but I would have had to fix them to the window framework, and I wouldn't be able to completely close the window once I stowed the air conditioner for the season. I went back to the store and bought some weatherstripping that I could attach to the panel instead, and it fit snugly. I then hooked up the air conditioner to the panel with a big flexible duct, and I was in business. The installation is more or less held in place by gravity, but it seems workable enough.

We had a cool spring, so it wasn't until June that the temperature was high enough to really make use of the air conditioner. It is noisy, but not impractically so, and a bit of a power hog, drawing over a kilowatt in air conditioner mode. I've been trying to be selective in its use, and had to tinker to figure out the optimum temperature setting -- not too cool, not too warm. It could chill me out as much as I wanted, however, so as long as the thing holds up, it seems like a pretty good buy.

I was thinking the air conditioner would end my restless summer nights, but it hasn't proven that simple. Given the noise, I would rather not run it after I go to bed if it's not really hot in my bedroom -- but I find that even if I'm slightly too warm, I can end up tossing and turning for hours. I've been tinkering with various procedures, but so far I haven't figured out the right formula. Nonetheless, I am happy with my new air conditioner. I still would have preferred a workable evaporative cooler -- but since there's no prospect of ever finding one, I just have to settle for this.



* EYE TRACKING ON THE PAD: In the past few years, gesture recognition has been in transformation from a sci-fi movie gimmick to practical reality. As reported by an article from THE ECONOMIST ("The Eyes Have It", 1 December 2012), eye tracking is the next frontier of human-machine interface.

The technology is already here; it's just that for the moment, it's not ready for the mass market. In Vivo BVA, a Paris-based marketing consultancy, has 19 mock supermarkets, known as "shopper labs", in Britain, China, France, Germany, Italy, and the USA. Test shoppers are given high-tech goggles that monitor their eye movements, and then set loose the shopper labs to see what they notice or don't notice. That allows In Vivo BVA to determine what logos, product catchphrases, and packagings catch the eye -- as well as "zombies", those that are subtly repellent.

Thousands of paraplegics are now also using eye trackers to control computers and wheelchairs. Eye tracking is also being used to alert drowsy drivers, diagnose brain trauma, train machine operators, and give surgeons with "a third hand" to control robotic equipment. The technology is still too expensive for consumer use, but costs are dropping rapidly. Haier, a Chinese manufacturer of household appliances, has developed an experimental TV set that uses eye tracking for control, and use of eye tracking by desktop computers, videogame consoles, and e-readers may not be far off.

Eye trackers combine a camera with an infra-red light source that illuminates the eye with bursts of invisible infra-red light. Some of this infra-red light disappears into the pupil, but some of it bounces back off the iris, the cornea, the eyelid, or the surrounding skin. An eye tracker maps the reflections from these elements elements and then performs an analysis to see where the eye is pointing.

Monitoring someone's gaze as they use a computer screen or watch TV is relatively easy; it can be done using an eye tracker aimed at the user from as far as two meters away. Tracking someone's gaze as they move around is trickier, demanding a special headset, with an eye tracker pointing towards the wearer's eyes, and an extra video camera facing forward to record what the wearer is seeing. The output from the eye tracker is then used to superimpose crosshairs or a colored dot on the video recorded by the camera, to show what the wearer was looking at.

A decade ago, headset eye trackers cost about $30,000 USD; today, they cost half that. Fixed eye trackers for computer users or TV viewers now cost about $5,000 USD; Tobii Technology, a Swedish company involved in user interface technologies, believes costs should drop to less than $3,000 USD in the near term.

The major application for eye tracking right now is in design and marketing, In Vivo BVA's work with test shoppers being a good example. The goggles are only part of the necessary technology; software sold by iMotions, a Danish firm, creates color-coded "heat maps" that shows exactly what test subjects find visually interesting, and what they find dull or annoying. Eye tracking works much better than simply asking test subjects what they like or don't like, people often finding it difficult to grasp what they actually feel and articulate it.

Companies like Haier are thinking of what eye tracking will be able to do a generation down the road; another big Chinese firm, computer maker Lenovo, has built a set of prototype laptop computers with Tobii eye trackers built into a bulge in the top of the lid. The Eye Tribe -- a Copenhagen-based startup spun off from the Gaze Group, a research center at the IT University of Copenhagen -- has modified a stock tablet computer to allow it to be controlled using eye movements; the firm is also developing eye-tracking software for Android, the leading smartphone operating system. Tens of thousands of users have downloaded the Android eye-tracking software.

Much of the tinkering right now is for the disabled, but there's no reason to stop there. Moving a cursor on a computer display is obviously much quicker and, in principle, easier with eye tracking than by using a mouse. Eye tracking with an e-reader could allow the reader to translate or define terms that aren't easily grasped; stare at phrase for a moment, the e-reader would highlight it. Toyota is developing an "attention monitoring" system for cars that will sound an alarm if a driver starts to fall asleep at the wheel. Mining and drilling companies are now obtaining drowsiness-detection systems for their big mining rigs. It's expensive, but half of all mining vehicle accidents are caused by fatigue, or possibly the occasional hangover, and even a minor accident is costly.

Of course the military is interested in eye tracking as well, with the glances of American pilots being monitored during flight simulations by an eye tracking system. The pilots find the trackers annoying because they catch every moment of inattention. Down the road, eye tracking could be used for targeting; right now, helmet-mounted sights are in common use, but during high-gee maneuvers a helmet can become so heavy as to make moving the head difficult or even hazardous. ISCAN, a defense firm in Woburn, Massachusetts, has developed a control system for a miniature attack drone in which the operator simply looks at the target, with the drone diving into it. Looks, it turns out, can kill.



* BITMESSAGE COVERS ITS TRACKS: As reported by BLOOMBERG BUSINESSWEEK, the recent flap over government data-mining of emails through its PRISM program has led to a surge of interest in secure online communications. One beneficiary has been Jonathan Warren, a New York-based programmer who was inspired by the Bitcoin digital currency scheme to create a secure email scheme he named "Bitmessage".

Bitmessage is based on a distributed encrypted network scheme in which all email is distributed to all users, with individual correspondence protected by encryption. Not only are the contents of emails kept secret, so are the sender and receiver, each using coded addresses based on a user's public encryption key. Send a message into the Bitmessage network, it goes to every node in the network, and only the recipient, having the private key, can pick it out: any one tree remains hidden in the forest. Messages are cycled out of the network after two days, recipients presumably having picked them out and saved them by then.

If a Bitmessage network gets too big, it is split into "substreams" to keep things manageable. Bitmessage also requires that a computation be performed for each message sent, raising the overhead for mass emailings. It doesn't handle attachments at present -- it seems possibly by intent, email attachments not being all that consistent with security.

Warren wrote Bitmessage in the Python programming language and put it on the internet for free download in 2012. Originally, most of the downloaders were Chinese, but since data-mining scandal broke, the bulk of download traffic switched to the USA. Right now, it's something of a toy for security-conscious tech geeks, some suggesting that Bitmessage will become much more popular once tacked onto popular email clients, such as MS Outlook and Google Gmail.

Others suggest that, at least in spirit, Bitmessage doesn't really mesh with conventional email. One might also suspect that when the current fuss over PRISM dies down, things will more or less go back to normal. Anyone with sense has known for a long time that email is painfully insecure; free encryption tools have been around for almost as long, if anyone thought it practical to use them. Still, Bitmessage is very slick, and it's good to have it in the tool rack. Once an idea has emerged, it is very difficult to make it go away.

* NSA DOES ANDROID: Also as reported by BLOOMBERG BUSINESSWEEK, the National Security Agency (NSA) -- America's electronic surveillance organization -- has provided code to Google to incorporate into Google's Android operating system. The NSA "Security Enhancements for Android" help to isolate apps and prevent hackers from spreading their tentacles through an Android system. That sounds far too much like the fox volunteering to guard the henhouse, with one blogger characterizing the reaction as: "ZOMG PRISM! THEY'RE BUILDING IN PRISM!!11ONE!!11ELEVEN!"

However, NSA has considerable expertise in data security, and there's nothing new about the security extensions: they were developed for the Linux open-source operating system over a decade ago. The security extensions for Linux are, as with the rest of the package, completely open source; they've been inspected by Linux experts not associated with the agency; nobody's seen that they do anything but what they're supposed to do; and it seems they do it quite well. Android being an offshoot of Linux, its security extensions will be open-source as well.

It makes perfect sense for the government to want a secure Android, since the government ends up using Android a great deal. On the other side of that coin, the NSA would have to be staggeringly clueless to try to pull off something sneaky and tell the public about it -- and nobody has accused them of being clueless.

In response to the discussion of the security enhancements, Apple publicly denied any involvement with the NSA, a company spokesperson saying: "Apple does not accept source code from any government agencies for any of our operating systems or other products." Of course not. Apple is dedicated to proprietary, and wants as little to do with open source as possible.



* MANUFACTURING ROBOTICS ENERGIZED (2): Robots are not just transforming factories. Another application is distribution -- robots zipping around automated warehouses at a sprint to store, retrieve, and pack goods for shipment.

C&S Wholesale Grocers, the nation's largest grocery distributor, has already deployed robot technology and envisions doing more. The company runs a huge distribution center at Newburgh, north of New York City, to support a major supermarket chain, with the facility contrasting the old and the new. The old system covers almost 50,000 square meters (a half million square feet) of the facility, with shelves loaded and unloaded 24 hours a day by hundreds of workers driving pallet jacks and forklifts. At peak times in the evening, the warehouse is a riot of beeping and darting electric vehicles, as operators with headsets are directed to cases of food by a computer that talks to them in four languages.

The new system, built by a Boston-area startup named Symbotic, is tucked into roughly 3,000 square meters (30,000 square feet) in a corner of the warehouse and is operated by a handful of technicians. The sub-facility is organized around a 21-level storage structure, with eight "rover" robots on each level, 168 rovers in all, shuttling through the pathways on each level at up to 40 KPH (25 MPH).

The rovers are linked to a central server via wireless, being given commands to access a particular item on their level; they shuttle to the item, grab it, then shuttle over to an elevator that sends the box down to the floor level. There the boxes are shunted onto a conveyor, with a robot arm taking the boxes and stacking them up in a 2.4 meter (8 foot) cube. The software is smart enough to pack the boxes in the proper order for convenient removal and shelving at the destination supermarket. Once the cube is finished, it is conveyed to machine that wraps in clear plastic for shipment, with a forklift operator summoned by the central server to pick it up and put it on the delivery truck. The system is extremely efficient and, thanks to its three-dimensional storage arrangement, very compact.

* Even robotics advocates admit there are plenty of things machines still can't do: construction jobs that require workers to move in unpredictable settings and perform different tasks that are not repetitive; assembly work that requires tactile feedback, such as fitting fiberglass panels into airplanes, boats or cars; and assembly jobs where only a limited quantity of products are made, or where there are many versions of each product, requiring expensive reprogramming of robots.

However, that list is growing shorter as the technology continues to advance. In a garage in an industrial neighborhood in Palo Alto, California, a robot being developed by Industrial Perception (IP) INC fitted with electronic eyes, along with a small scoop and suction cups repeatedly picks up boxes and drops them onto a conveyor belt. It is doing what low-wage workers do every day around the world. It seems like a relatively easy task for a robot, but the unpredictability of configuration and placement of the boxes made it difficult, vision systems being expensive and limited.

Microsoft broke the barrier for vision systems with the Kinect motion-sensing unit for the Xbox game system, Kinect providing high capability at an unprecedentedly low cost. A robot smart enough to see and pick up packages is extremely attractive to firms such as Federal Express and United Parcel Service, which now employ tens of thousands of workers doing such tasks.

IP is the first spinoff of Willow Garage, a robotics research firm based in nearby Menlo Park. IP is now trying to sell their tech to a company that currently employs thousands of workers to load and unload trucks. The workers can move one box every six second on the average, but the boxes can be heavy, tiring the workers and sometimes causing injuries. IP will win the contract if their robot can move one box every six seconds, and the firm's officials are confident they can meet that goal. [TO BE CONTINUED]


[FRI 19 JUL 13] WASTE NOT (10)

* WASTE NOT (10): The tale of the Fukushima nuclear accident cleanup leads to the bigger question of radioactive waste disposal in general. As reported by THE ECONOMIST ("Hot Stuff", 2 June 2012), around 270,000 tonnes (297,000 tons) of high-level radioactive waste, mostly spent fuel, are in temporary storage around the world, with another 10,000 tonnes (11,000 tons) added every year.

What to do with it? Bury it in an undersea fissure? Store it underground? Process it into a less hazardous forms? The question seems particularly open right now, since in 2009 the US abandoned plans to store nuclear waste in a huge storage site shielded by 300 meters (1,000 feet) of volcanic rock in Yucca Mountain, Nevada. A presidentially mandated Blue Ribbon Commission on America's Nuclear Future, which delivered its final report in January 2012, emphasized an "urgent" need for a new waste-disposal strategy. The European Union is also investigating the matter, new rules drawn up in 2011 requiring member countries to draw up long-term plans for dealing with their nuclear waste by 2015.

The big, obvious problem with handling nuclear waste is time. Not only does the waste have to be put away securely, it has to stay secure for historical timeframes, even to future eras where people might not know what a nuclear waste storage site is all about, and think it would be interesting to crack into it. Three basic approaches are being considered:

The hottest power-station waste, high-level spent fuel, is usually cooled in pools of water for at least five years and then packed away in thick concrete-and-steel "dry casks", each the size of a small truck. A cask can hold out for a few hundred years, maybe even a few thousand -- but there's no way a dry cask would last 200,000 years, which is roughly how long it takes for the longest-lived isotopes of plutonium to decay to harmlessness.

Storage and handling technology is being refined, however. New materials for casks are being developed, one candidate being "Alloy 22", composed of nickel with chromium for corrosion resistance, with molybdenum and tungsten for strength. The Australian Nuclear Science & Technology Organization has devised a process using heat and high pressure to make synthetic rock out of radioactive waste and minerals including calcium, titanium and zirconium. Called SynrocANSTRO, it reduces radiation by "locking it up safely in minerals", according to a spokesman. America's Department of Energy (DOE) wants to use the process to dispose of radioactive powder in Idaho left over from the extraction, decades ago, of uranium from spent fuel.

Given emerging improvements, why not just wait for better solutions to become available? Unfortunately, that would simply pass the problem down to future generations, and the US Blue Ribbon Commission commented in a report that there is "no ethical basis" for doing that. Besides, do we really just want the waste sitting around more or less unsecured indefinitely? It would be an accident waiting for a time to happen, with natural disasters or conflicts possibly disturbing temporary storage solutions. The general consensus is to try to deal with the problem in the here and now.

But how? Dumping it into a volcano would be dodgy to the point of absurd, and though shooting it into the Sun would certainly be a thorough solution, it would be dodgy as well -- due to the threat of launch accidents -- and insanely expensive. One idea floated early on in the nuclear age was to bury waste canisters in Antarctic ice, with the canisters then gradually melting themselves down to the bottom; there were some difficulties with the concept, and it was given up in the 1970s. At that time a few countries, including France, dumped small quantities of low-level radioactive waste into the deep ocean; that scheme that worked reasonably well for the amounts involved, but such dumping was banned in 1993.

The idea has been reconsidered, current thinking suggest that burying nuclear waste in the deep sea bed might be practical. Deep-sea drilling is a mature technology, having been refined by decades of oceanic oil exploitation, and it would be technically practical to bury canisters of radioactive waste in areas lacking volcanic activity, placing the canisters in holes plugged up with seabed clay. One notion associated with this concept is to bury radioactive waste in a "subduction fault", where one tectonic plate slides below another into the Earth's mantle, carrying the waste along with it.

That's an interesting idea, but the US gave up on seabed disposal studies in 1986 -- it appears over potential clashes with international environmental law -- with other countries then giving up on it as well. However, there is still work along such lines; a California startup named Permanent RadWaste Solutions has patented a steel, copper and lead container shaped to withstand pressure in a subduction fault. It looks like a bowling pin the size of a car, and is designed to close up ever more tightly on itself as pressure increases. Subduction-zone disposal has its attractions, one being that it is effectively in nobody's back yard, but there's no optimism of any such thing happening for the time being. [TO BE CONTINUED]



* Space launches for June included:

-- 03 JUN 2013 / SES 6 -- A Proton M Breeze M booster was launched from Baikonur in Kazakhstan to put the "SES 6" geostationary comsat into orbit for SES of Luxembourg. The spacecraft was built by EADS Astrium and was based on the Astrium E3000 comsat bus. It had a launch mass of 6,100 kilograms (13,450 pounds), a payload of 43 C-band / 48 Ku-band transponders, and a design life of 15 years. SES 6 was placed in the geostationary slot at 40.5 degrees West longitude to provide communications services to North America, Latin America, the Caribbean, Europe, and the Atlantic region. It replaced the aging NSS 806 comsat at the location, NSS 806 having been launched as Intelsat 806 in 1998.

-- 05 JUN 2013 / ATV 4 -- An Ariane 5 ES booster was launched from Kourou in French Guiana at 2152 GMT to put the fourth ESA "Automatic Transfer Vehicle (ATV 4)" unmanned cargo spacecraft, named "ALBERT EINSTEIN", into space on an International Space Station (ISS) support mission. It carried a payload of 6.5 tonnes (7.15 tons). It docked with the aft port of the Russian Zvezda ISS service module ten days after launch.

ATV 4 approaches ISS

-- 07 JUN 2013 / COSMOS 2486 (PERSONA) -- A Soyuz 2-1b (Fregat) booster was launched from Plesetsk Northern Cosmodrome in Russia at 1837 GMT on 7 June to put a secret military payload into orbit. The spacecraft was designated "Cosmos 2486" and was believed to be the second Persona spy satellite. The first was launched in 2008, but failed shortly after arrival in orbit.

-- 11 JUN 2013 / SHENZHOU 10 -- A Long March 2F booster was launched from Jiuquan in northern China at 0938 GMT to put the "Shenzhou 10" manned space capsule into orbit, delivering three Chinese taikonauts to the Tiangong 1 space station, launched on 29 September 2011. The taikonauts included mission commander Nie Haisheng -- the only member of the crew to have flown before, on the 2005 Shenzhou 6 mission -- Zhang Xiaoguang, and Wang Yaping, Wang being the second female taikonaut. The crew returned to Earth in Inner Mongolia on 26 June after 15 days in space. This was the fifth Chinese crewed space mission.

-- 25 JUN 2013 / RESURS P1 -- A Soyuz booster was launched from Baikonur at 1728 GMT to put the "Resurs P1" civil Earth resources observation satellite into Sun-synchronous orbit. It was the first launch of the new Resurs P series, replacing the Resurs DK series, with the Resurs P carrying a hyperspectral imaging payload, a wide-angle imager, and a high-resolution imager.

-- 25 JUN 2013 / O3B x 4 -- A Soyuz 2-1b (Fregat) booster was launched from Kourou at 1927 GMT to put four "O3b" comsats into equatorial medium Earth orbit for O3B Networks LTD. "OTB" stood for "Other Three Billion", reflecting the comsat constellation's mission of providing broadband communication access to undeveloped countries. The O3B satellites were built by Thales Alenia Space, had a launch mass of 700 kilograms (1,545 pounds), and a payload of ten Ka-band transponders coupled to twelve antennas. Two more batches of four satellites each were to be launched to complete the constellation.

O3b satellite

-- 27 JUN 2013 / IRIS -- An Orbital Sciences Pegasus XL booster was launched at 1427 GMT from the Orbital Tristar Stargazer carrier aircraft operating out of Vandenberg AFB in California to put the NASA "Interface Region Imaging Spectrograph (IRIS)" satellite into orbit. The spacecraft was a "Small Explorer" space science satellite, with a launch mass of 140 kilograms (300 pounds) and a 20 centimeter (8 inch) ultraviolet telescope feeding a multichannel spectrograph. The instrument had very high resolution in a narrow field of view; it was to be used to perform observations of energy transfers through the Sun's atmosphere and heliosphere.



* CLEVER FIGARO: The parrot family -- including parrots, cockatoos, and parakeets -- is noted for its intelligence, some species such as the African gray being able to hold simple conversations with their human keepers. As reported by AAAS SCIENCE NOW ("The Innovative Cockatoo" by Virginia Morrell, 5 November 2012), a Goffin's cockatoo named Figaro -- a member of a captive colony in Austria, the birds being originally native to Indonesia -- has now demonstrated the ability to used little stick tools.

Other birds, most notably the New Caledonian crow, have demonstrated tool-using capabilities, but Figaro is different: while New Caledonian crows will use tools in the wild, nobody has ever seen cockatoos do it in their native environment. Figaro's unusual talent came to light when a student observer at the University of Vienna's Goffin lab noticed the cockatoo playing with a pebble. The bird dropped the stone through the aviary's wire mesh, where it landed on a wooden beam. Figaro tried to retrieve the stone with his claw. When that failed, he picked up a piece of bamboo and used it to try, unsuccessfully, to rake the stone back into the aviary.

The student reported that to Professor Alice Auersperg; Figaro was promptly segregated into another aviary, with just female cockatoo named Heidi to keep him company. They isolated him because they were afraid that his fellow cockatoos would get ideas from him, confounding studies. They set up a video camera to observe him, and then set a nut on a wooden ledge outside the aviary cage, just out of Figaro's reach. First he found a little stick on the aviary floor, but it wouldn't do the job; he then decided to tear a large splinter off the interior side of the ledge, trim it appropriately, and use it to rake in the nut.

The researchers gave Figaro nine other tests, and he aced them all except one, Auersperg saying: "He did everything: tool use, tool manufacture, and tool modification, and he made them so quickly. His second tool took him less than five minutes to make, which is a drastic reduction." Other cockatoos tested individually didn't quite get the tool-making concept; it is unclear if the researchers then let Figaro train the others in the matter, which would have been an interesting test in itself.

The experiment suggested that tool-using doesn't necessarily require brain specialization, it just requires brains: if an animal's clever, it can be clever enough to make tools. John Marzluff, a wildlife specialist on crow cognition at the University of Washington, says that parrots in general "have big brains, are social, and long-lived -- that foster innovative, cognitive problem solving." He adds that with plenty of food around in the wild, parrots don't need to be inventive: "But put that same mind behind bars and give it a reason to make a tool, and bingo, a tool is made. It makes you wonder what other thoughts are rattling around the brains of pet parrots everywhere."

ED: What other thoughts indeed, this article bringing back memories of Snowball the crested cockatoo, who became a YouTube star some years ago by dancing to the heavy beat on video. He bobbed his head, raised one leg and then another, and kept time with the music even when the time was varied. It's of course impossible to figure out what's going on in an animal's head, but anyone watching could only conclude that Snowball really liked to "get down".

Along similar lines, WIRED Online reported that Andrew Gray, an engineering student at the University of Florida, had a problem: his pet African gray parrot, Pepper, would start screaming when got to feeling bored or lonesome -- and he could scream very loud. At first, Gray tried nailing Pepper with a shot of water from a squirt bottle, which seem to work for a while. That being laborious, Gray then set up an automated system that would squirt Pepper every time the bird screamed. No joy; Pepper started screaming just to get a bird bath.

Pepper's bird buggy

Gray finally realized that Pepper's difficulty was that he didn't like being alone -- parrots are sociable creatures -- but his wings being clipped, he couldn't easily go to another room to find human company. Gray decided to give him mobility, giving him a "bird buggy", which was a flat blue box with wheels, a perch, and a joystick on a stand in front of the perch. Pepper can sit on the perch and manipulate the joystick with his beak. The bird buggy also has infrared and bump sensors to deal with running into things, and even a homing system to allow it to park and recharge itself when Pepper isn't using it. Gray put together a video that suggests Pepper was a fair driver.



* DO YOU KNOW WHERE YOUR KIDS ARE? Thanks to modern mobile communications tech, it is now easy to stay in touch with one's kids. As discussed by an article from THE ECONOMIST ("Chips Off The Old Block", 12 January 2013), there is a fine line between that and wanting to keep an electronic leash on them.

Technology for tracking kids have been around for about a decade. In some cases, it's hard to object -- parents of autistic or otherwise challenged kids, who can easily get lost or into danger, have good reason to want to know where their children are. Similarly, youngsters on some Canadian farms wear radio tags so their farmer parents know where the kids are while driving a combine or tractor around.

It is simple enough just to give kids smartphones and track them with that. Mobile operators often offer child-tracking services at extra cost, but now parents can just download the free app "Life360" instead. Life360 had a million registered users in 2010, tens of millions now; projections are that over 100 million Americans and Europeans will be tracking family members by 2016. Such services can give children's location; send alerts about their behavior; log when they return home, go outside a specified boundary, or go out late. They can even determine if a child is in a vehicle and if the vehicle is speeding; one product disables a phone's text messaging capability in a moving car. Life360 highlights the addresses of sex offenders.

Parents in Japan and America are the keenest on such technology. Europeans are apparently more relaxed about child safety and have tougher privacy laws, with some European countries mandating that such tracking involve the minor's consent. However, families in South America, worried about gang crime and kidnapping, are very interested in the technology.

Of course public authorities like tracking, too. Schools in Osaka, Japan, began issuing radio frequency identification (RFID) tags to students in 2004. Sensors in school buildings read them to check pupils' attendance and location, though the kids are not tracked when they leave the school. In Dubai, the same technology notifies parents when their children get on or off school buses. In 2012, two schools in San Antonio, Texas, embedded RFID tags in identity badges belonging to their 4,200 students. The tags permit counting of truants; since funding is linked to daily attendance, the system allows schools to claim more taxpayer cash. One girl in the school system refused to wear the badge, saying it was the "Mark Of The Beast", some religious puritans believing RFID is a sign from the Book of Revelations; the case has been working its way through the courts, the courts being sympathetic to the school system so far.

Identifying RFID with Satan may seem to be taking things more than a bit too far, but obviously tracking kids presents hazards. Some don't see much benefit in it, pointing out that kidnappers will of course ditch a kid's phone on the spot, and tracking won't really do much to protect kids from accidents. Others fear that children who submit to tracking in schools will more easily accept state surveillance in adulthood.

There is a general perception that tracking does have benefits and is here to stay, suggestions being floated to make it more palatable. One is for schools to make tagging voluntary and to make sure sensors are visible. Another fix for kids with smartphones is to allow them to push a button to tell their parents where they are -- that is, they are tracked to the extent that they want to be. There are also apps that allow users who feel at risk to send a message with a location to a list of trusted contacts. It will take a generation for people to get used to tracking services; in doing so, they will gradually converge on a solution they are comfortable with.



* MANUFACTURING ROBOTICS ENERGIZED (1): There's been a lot of talk of the impact of 3D printing and other new technologies on the automation of manufacturing; as reported by an article in THE NEW YORK TIMES ("Skilled Work, Without the Worker" by John Markoff, 18 August 2012), somewhat lost in discussion is the impact of 21st-century robotics in the field.

Welcome to the Philips Electronics factory at Drachten in the Dutch countryside, where a set of 128 robot arms assembles electric shavers with a dexterity beyond the capability of most humans. They work at dizzying speed, all day and night, every day of the year. Contrast this to the Philips factory in the Chinese City of Zhuhai, which has hundreds of workers putting together shavers. The Drachten plant only requires a tenth of the headcount to operate.

Royal Philips Electronics began making the first electric shavers in 1939 and set up the factory in Drachten in 1950. Binne Visser, the assembly line boss in Drachten, takes pride in the sophistication of the latest shavers. They sell for as much as $350 USD and, he says, are more complex to make than smartphones. The assembly line is made up of dozens of glass cages housing robots made by Adept Technology. Video cameras on top of the cages guide the robot arms to pick up the parts they assemble; the arms bend wires with millimetric accuracy, set toothpick-thin spindles in tiny holes, grab miniature plastic gears and set them in housings, and snap pieces of plastic into place. The robot factory technology is not just limited to assembling shavers either, with Visser commenting: "With these machines, we can make any consumer device in the world."

Change scene to Milpitas, California, south of San Francisco, where the Flextronics company operates an enormous solar-panel factory. All of the heavy lifting and almost all of the precise work is done by robots that piece together solar cells and seal them under glass. The relative handful of human workers do things like trimming excess material, threading wires and screwing a handful of fasteners into a simple frame for each panel. Flextronics does not hire hordes of workers from the local community, but that wouldn't work anyway: a traditional factory set up in the USA to turn out solar panels couldn't meet Chinese competition, all the more so because the Chinese are increasingly turning to industrial robotics themselves.

Industrial robots have been around for more than a generation, but traditionally they were only useful for repetitive and brute-force jobs. Anything that required dexterity to assemble required human workers. Machine intelligence has advanced rapidly in the past few decades, with many thinking the days of farming out manufacturing to nations with low-cost work forces will soon be a thing of the past. Manufacturers would prefer to build product in the markets where the products will be sold, reducing logistics costs and worries about intellectual property theft -- as well as worries about bad publicity over subcontractors who abuse their employees. There are no rights advocates for robots, at least not yet.

There are skeptics who think that robots won't be able to do everything in the near future. Bran Ferren, a roboticist at Applied Minds in Glendale, California, commented: "I had an early naivete about universal robots that could just do anything. You have to have people around anyway. And people are pretty good at figuring out, how do I wiggle the radiator in or slip the hose on? And these things are still hard for robots to do."

* Still, robots can do more and more all the time. Rapid improvement in vision and touch technologies is putting more and more manual jobs within the abilities of robots. For example, Boeing's wide-body commercial jets are now riveted automatically by giant machines that move rapidly and precisely over the skin of the planes.

The latest manufacturing robots is more flexible and easier to train. Electric-car manufacturer Tesla Motors is now working on the leading edge of robotics in the company's factory in Fremont, California, on the edge of Silicon Valley. The factory used to turn out Toyota Corollas for the US market; half of the plant is still shuttered and dark. The other half is brilliantly lit, full of big, intimidating-looking red robot arms.

Tesla loves robots

While robots have long been used in auto production, the robots in the Fremont plant are different, able to change their "hand" to do different tasks. Traditional robots could only do one task, but the robots in the Tesla factory can do four: welding, riveting, bonding, and installing a component. Up to eight robots work rapidly around a car as it stops at each station along the line. The projection is for the production of over 80 cars a day. Tesla is planning to introduce an SUV soon, and it will be produced on the same line, the robots just using different software.

Tesla's factory is small by auto industry standards. Hyundai and Beijing Motors recently completed a mammoth factory outside Beijing that can produce a staggering million vehicles a year, using more robots and fewer people than earlier plants. The plant will feature the same sort of flexible manufacturing robots as those at the Tesla plant. [TO BE CONTINUED]


[FRI 12 JUL 13] WASTE NOT (9)

* WASTE NOT (9): As a final topic on the tour of waste management, there is the issue of its ultimate challenge: nuclear waste. The earthquake and tsunami that hit Japan in 2011, discussed here at the time, crippled the nuclear facility at Fukushima and spread radioactive contamination around the surrounding prefecture. As reported by an article from AAAS SCIENCE ("Cooling A Hot Zone" by Dennis Normile, 1 March 2013), the Japanese government is now grappling with the cleanup problem.

The Fukushima accident was not as disastrous as the 1986 Chornobyl meltdown, only releasing (very roughly) a third the level of radioactive contaminants. In addition, while long-lived radionuclides such as isotopes of plutonium figured heavily in Chornobyl fallout, the fallout in Japan was dominated by cesium-137, with a half-life of only 30 years. About 80% of that fallout went out to sea, where it was dispersed; there are areas still off-limits to fishing, but the expectation is that problem will fix itself in time as the cesium-137 decays.

However, that still leaves the Japanese Environment Ministry confronted with the challenge of cleaning up Fukushima prefecture itself, handling anywhere from 16 million to 41 million cubic meters (21 to 54 million cubic yards) of contaminated waste. The region has been mapped into five zones, graded by the intensity of the radiation. The "hottest" zone, roughly in a radius of 20 kilometers (12 miles) of the crippled reactor, is off-limits indefinitely; naturally, the government is focusing on the "cooler" areas in hopes of allowing citizens to return to their homes there over the short term.

The first cleanup step is removing contaminated topsoil, pavement, and leaves, as well as washing down building surfaces. The Japan Atomic Energy Agency (JAEA) has evaluated decontamination procedures and produced guidelines on how to do the job. The decontamination is piling up mountains of low-level radioactive waste, in addition to the hundreds of thousands of garbage bags littering the area, containing waste collected by citizens and municipal workers.

The government wants to move all this waste into interim storage sites -- landfills lined with double walls and floors of concrete surrounded by bentonite, a claylike sealant, and covered by topsoil. However, building such storage sites is expensive, one big factor being that land is somewhat at a premium in Japan, and the JAEA has been funding research on means of reducing the volume of waste. Construction and chemical companies, university research groups and government agencies -- even the Japan Aerospace Exploration Agency (JAXA), the Japanese space agency -- have been coming forward with proposals. Many ideas seem promising, with the government evaluating them on the basis of effectiveness versus cost. Ultimately, however, the mess has got to be cleaned up, and the government will have to live with the cost. [TO BE CONTINUED]



* SCIENCE NOTES: As reported by THE ECONOMIST, caffeine is toxic to plants, and so coffee grounds cannot be disposed of carelessly. Some process for decaffeinating the coffee grounds would be handy since they are otherwise nutritious, and could be used as a component of animal feed.

There has been work on using bacteria to do the job, originally focusing on a species named Pseudomonas putida -- but it was inefficient, only able to deal with small quantities of coffee grounds. Now researchers at the University of Texas in Austin have spliced the genes from P. putida for the enzymes that digest coffee grounds into Escherichia coli -- the human colon bacterium, which has a long history as a lab guinea pig and in biotech. The hybrid bacterium didn't work all that well at first, but the researchers persevered, finally splicing in a small component from a third bacterium, Jathinobacterium. The chimera proved highly effective; the research team is now trying to scale up to a production process.

* In slightly related news, as discussed by THE ECONOMIST, Americans are literally addicted to caffeine, now available in a wide range of tasty foods and beverages. However, given such a massive national caffeine rush, the question has come up of just how safe all that caffeine is. America's Food & Drug Administration (FDA) has now announced it will study the impact of caffeine; the city of San Francisco has taken a more aggressive approach, suing Monster Beverage, a Californian firm, for marketing high-caffeine drinks to kids.

US sales of energy drinks were estimated as $8.6 billion USD in 2012, a dozen times their level a decade earlier. Red Bull is the leader in America with 36% of sales, while Monster has 29%, and Rockstar 8%. Monster Energy and Rockstar are sold in big cans, and have 160 milligrams (mg) of caffeine; for comparison, a standard can of cola has only 35 mg. The FDA says adults should consume no more than 400 mg a day; more than that can increase anxiety and blood pressure. The number of emergency hospital visits in the USA related to energy drinks doubled between 2007 and 2011, to more than 20,000.

Too much caffeine?

So far, caffeinated products have been lightly regulated; the FDA's only explicit rule for caffeine dates from the 1950s and applies only to cola. However, in 2010 the FDA warned that adding caffeine to alcoholic drinks promoted "risky behavior", for the simple reason that drunks are likely to do more stupid things as long as they stay awake. The FDA's new action has given the energy drink business the jitters, so to speak, with Monster's sales declining significantly in response. Picking on energy drinks does come across as a bit unfair, since a large coffee from Starbucks has more than twice as much caffeine as a can of Monster -- but nobody's daring to take on a nice cup of coffee in the morning just yet.

* As reported by WIRED Online blogs, space astronomy missions such as the NASA Kepler telescope have proven very productive in spotting candidate planets in distant star systems, by the simple measure of maintaining a watch on a star and observing it for slight motions or transient variations in brightness.

Now an array of four 70-centimeter (28-inch) automated telescopes is being set up on Mount Hopkins in Arizona to continue the exoplanet search from the ground. Each night, the telescopes of the "Minerva" array will zero in on a target star within 75 light-years of Earth and watch it carefully. The research team hopes to find about a dozen new planets, most of them two to three times the size of Earth and a few of them orbiting in the habitable zone where liquid water could exist. The project will also find Earth-sized or smaller planets, but these will likely orbit too close to their parent star to host life.

The four telescopes will be commercial off-the-shelf items, with a cost of about $225,000 USD each. They will be fitted with a custom-built spectrometer; along with the telescope shelter and the control system, Minerva will cost about $3.5 million USD. That may sound like a lot, but Kepler cost $600 million USD, and it was not regarded as a high-budget space project.



* WE'RE THE GOVERNMENT, WE'RE HERE TO TWEET: As reported by an article from BLOOMBERG BUSINESSWEEK ("The Government Learns To #lovetwitter" by Claire Suddath, 20 May 2013), in 2009 US President Obama issued the "Open Government Directive", requiring every government agency to make its reports easily available online to the public. In response, they've jumped enthusiastically into social media.

Take, for example, the Department of the Interior, stereotyped as a gray government bureaucracy. On 4 May 2013, STAR WARS Day, Interior sent out tweets linking to behind-the-scenes pictures of the filming of the movie RETURN OF THE JEDI on the Imperial Sand Dunes in California, announcing: "Did you know that RETURN OF THE JEDI was filmed on [public] lands?" Interior is fond of tweeting out photos of baby animals to its more than 70,000 followers.

The White House is the King Kong of USGOV twitterers, with 3.8 million followers, who are sent updates on White House activities, as well as official photos and snaps of Bo, the Obama dog, sent out by Michelle Obama. Says an Interior staffer in charge of digital communications: "The White House is the best at this. Most of us are still trying to figure stuff out."

tweets for America

Each Federal agency conducts its own social media program. They don't get any additional money for it, but most have long had media offices, and extending their activities to tweeting is not troublesome. Different agencies tend to have different group of followers; for example, the US General Accounting Office (GAO), which audits the finances of government programs, doesn't have a big public following. Says a GAO official: "We're not that popular on Twitter, but that's because our audience is Congress."

Indeed, the tweets of most Federal agencies tend to be businesslike, colorful tweets not always being appropriate to the mission. As a Pentagon official put it: "If the Department Of Defense is lighthearted, that doesn't bring a tremendous amount of comfort." Putting a happy face on America's military machine just doesn't work. However, the US Census Bureau, which might be seen as stodgy, has proven surprisingly popular, with over 38,000 followers, due to the bureau's interesting feed of geeky facts. For example, on President's Day the Census Bureau sent out tweets about the number of places named after US presidents; in 2012, the bureau collaborated with the National Archives to send out a series of tweets about the 1940 census. However, the viral moment was in August 2012, when the bureau announced that the US population had reached 314,159,265 -- the same as the first nine digits of pi -- with the announcement retweeted a hundred times. It might rightly be commemorated; a tenth digit is not likely to be added any time soon.



* HAULING OIL: As reported by an article from BLOOMBERG BUSINESSWEEK ("All Aboard The Crude Express" by Matthew Philips & Asjylyn Loder, 17 June 2013), America's recent oil production boom has been good for business -- one particular beneficiary being the railroads.

The railroads are now hauling more crude than at any time since World War II. In the first quarter of 2013, trains pulled a record 97,135 carloads of crude oil -- 166% more than in the first quarter of 2012, and over nine times as much as was transported in all of 2008. While moving crude by pipeline costs only a half to a third as much as carrying it by rail, given high oil prices the transport cost is not that big a pinch, while it takes both time and a lot of money to build a pipeline. Pipelines are more inflexible as well; it is less effort to build a spur line and a terminal to link the current rail network to an oilfield.

oil train, Montana

That flexibility means that oil producers can shift deliveries to a region where prices are highest, and with a lot of oil being shipped, the relative expense of doing it by rail is not a big deal. Oil companies are frantic to lease rail cars, with the big rail firms ramping up their capacity in pace.

Much of the boom at the moment is from the Bakken shale formation in North Dakota, where crude production has jumped about 250% since 2009. The BNSF and Canadian Pacific railroads have been investing heavily in the site, and now haul off about 400,000 barrels from it each day -- incidentally, a single train can handle about 70,000 barrels of oil. The oil boom has helped the railroads weather the fall in coal shipments, as powerplants shift to clean natural gas. Oil fetching a better price than coal, the railroads can also make more profit off oil than coal.

Of course, over the long term pipelines really are more economical, and they are gradually being implemented to handle "main trunk" routes. The industry expects that the mad rush for more train capacity will peak, more likely sooner than later -- but pipelines will never be able to match the reach of the rails, and even pipeline companies are investing in rail haulage. The freight train, basically a 19th-century technology, still remains entirely relevant in the 21st century.

* ED: This article puts an interesting spin on the current controversy concerning the Keystone XL pipeline project, intended to hook the oilfields of Alberta to refineries in Nebraska. The Obama Administration has been waffling on the issue, torn between environmental and economic concerns -- the economic concerns enhanced by the desire to keep the Canadians happy. The interesting fact is that if the oil doesn't go by pipeline, it will then go by train, and will anyone be better off?

This question was underlined on 6 July, when an oil train derailed and then brewed up in the Quebec town of Lac-Megantic, causing massive destruction and leaving about 50 citizens dead or missing. I cannot say what the answer to that question is, but I can say there is definitely a question there.



* NOT SO TOUCHY (2): To obtain a better understanding of why the Earth's temperature has leveled off over the last decade, we need to have a better understanding of the factors that affect global climate beyond a simple increase in CO2. For starters, consider aerosols -- such as sulphates in the atmosphere. They reflect sunlight, reducing warming; but some aerosols increase warming as well. On the balance, the consensus is that they cause a cooling effect, of up to half a degree Celsius. Possibly aerosols might account for the leveling off of temperature?

Possibly not. Over the past few years years, measurements of aerosols have improved enormously, with detailed data from satellites and balloons suggesting their cooling effect is lower than previously thought -- and where applicable, their warming effect is greater. Soot is a significant aerosol in terms of warming because it absorbs sunlight; recent studies have enhanced its impact on warming. That means it's unlikely that aerosols are responsible for the leveling off of temperature.

What else might be causing it? Clouds are another factor to consider, and though they typically cause warming in models, they are notably tricky -- an unreleased IPCC report saying that "the cloud feedback remains the most uncertain radiative feedback in climate models." They can trap heat, but they also reflect sunlight, and may have other effects.

Another wild card in the climate game is the effect of oceans. Over the past decade, the long-term rise in surface seawater temperatures seems to have stalled; that suggests that the oceans are not holding back the temperature rise by soaking up heat from the atmosphere. As with aerosols, this conclusion is based on better data from new measuring systems -- but it only really applies to surface temperatures, the top 700 meters (2,300 feet). Studies hint that the deeper ocean has continued to warm, and so that might lead to the temperature plateau.

Finally there is the possibility -- one which climate change denialists have made much of -- that natural climate variability has skewed perceptions of the significance of man-made warming. A paper by Ka-Kit Tung and Jiansong Zhou in the PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES linked temperature changes from 1750 to natural changes, such as sea temperatures in the Atlantic Ocean, and suggested that "the anthropogenic global-warming trends might have been overestimated by a factor of two in the second half of the 20th century." Could the rise in temperatures in the 1990s and the flattening in the 2000s have been caused in part by natural variability?

Maybe, but climate denialists shouldn't get too smug; for all we know, the recent flattening may be due to a natural cycle working against warming -- and once warming starts to rise again, it could then do so with a vengeance. On what basis of knowledge do the skeptics judge which is the case? Reto Knutti of the Institute for Atmospheric and Climate Science in Zurich, who generated an EBM giving relatively modest warning, cautions that "the bottom line is that there are several lines of evidence, where the observed trends are pushing down, whereas the models are pushing up, so my personal view is that the overall assessment hasn't changed much."

In other words, we have valid theoretical reasons to take global warming seriously, but observations suggest we're missing something. Still, given that observed realities are proving less alarming than expected, a small reduction in estimates of climate sensitivity would seem justified. If climate scientists were credit-rating agencies, climate sensitivity would be on negative watch -- but it would not yet be downgraded.

Climate scientists do not believe the issue of climate change has gone away. As a rule of thumb, global temperatures should rise by about 1.5 degrees Celsius for each trillion tonnes of carbon put into the atmosphere. The world has pumped out half a trillion tonnes of carbon since 1750, and temperatures have risen by 0.8 degrees Celsius. At current rates, the next half-trillion tonnes will be emitted by 2045, the one after that before 2080. Since sinks for CO2 work slowly, these human inputs accumulate, and the result could be an increase in temperatures compared with pre-industrial levels by around 2 degrees Celsius even with a lower sensitivity, maybe double that at the high range of estimates.

Climatology is admittedly an uncertain sort of science; but given that uncertainty, and our ignorance of the possible effects of climate change, dismissing the issue as a delusion or a conspiracy is carelessness. We would greatly prefer to believe the optimistic forecast if it's really justified by the facts; if, in the end, honest research -- as opposed to blind denial -- shows worries to have been exaggerated, whatever embarrassment felt by the climate research community from overstating the case will be well less significant than the relief at the realization that the case isn't as dire as feared. [END OF SERIES]


[FRI 05 JUL 13] WASTE NOT (8)

* WASTE NOT (8): Another article in the AAAS SCIENCE series on waste management ("Getting Minds Out Of The Sewer" by Greg Miller) followed up the article on the Blue Plains wastewater facility by pointing out that wastewater plants produce another useful resource: clean water. The world is beginning to suffer from a clean water crunch, and any useful source of the stuff is all for the good. From a technical point of view, it's not all that hard to do.

The hard part is to get people to accepting the idea of drinking water extracted from sewage. Orange County, California, has limited resources of clean water; the county opened a facility in 2008 that can produce 270 million liters (70 million US gallons) of recycled water a day, enough to serve 600,000 residents. However, instead of pumping the water to residents, the county pumps it into the ground -- half to prevent Pacific seawater from leaking into aquifers, half into basins where it filters through sand and gravel to replenish local aquifers. Citizens don't like the idea of drinking the recycled water directly, but they don't worry if it's been pumped into an aquifer, then pumped out again for drinking: out of sight, out of mind. The facility is named the "Groundwater Replenishment System" as a bit of sleight of hand to help with the dissociation.

A recent report from the National Research Council of the US National Academies of Sciences claimed that wastewater reuse could provide over a quarter of the public water supply in US coastal communities, if the instinctive disgust factor can be dealt with. That instinctive response isn't always reasonable, as demonstrated in experiments conducted by Paul Rozin, a psychologist at the University of Pennsylvania, who has investigated human disgust reactions. In one experiment in the 1980s, he gave students a piece of fudge shaped like a dog turd, Rozin saying: "They know it's chocolate, okay, and they like chocolate, but most of them won't eat it."

Another experiment by Rozin and his colleagues showed just how far-reaching that aura of disgust is. A researcher showed students a glass of juice, and then dipped a dead, sterilized cockroach into it. Despite assurances that the juice was still perfectly clean, the students wouldn't drink it. That wasn't too surprising, but if the researcher then poured juice into other glasses that hadn't been contaminated, students still didn't want to drink it. That sense of aversion has helped torpedo wastewater recycling projects in San Diego, Los Angeles, and elsewhere. Paul Slovic -- a psychologist from the University of Oregon in Eugene and president of a nonprofit research organization named Decision Research -- says that opponents have used slogans such as "toilet to tap" to generate public revulsion and resistance to such efforts.

Since the 1990s, Slovic has study the psychology of risk assessment, observing that while the experts trot out lists of benefits versus risks, people typically rely on their intuition to make a judgement, and then select facts as guided by that judgement. Slovic says: "For most of us, risk perception is not the output of a scientific, mathematical calculation, but of a gut feeling."

Water agencies now understand the challenge and have been funding studies of their own. A 2008 survey of public attitudes sponsored by WateReuse -- a nonprofit organization in Alexandria, Virginia -- that involved Rozin, Slovic, and others found that citizens were more willing to accept water reuse when it was explained to them that wastewater facilities usually dump their clean water back into a river, and so communities downstream were drinking recycled water anyway. A further study suggested that it was best to publicly emphasize the cleanliness of the water, talk as little as possible over its origins, and get people to think in terms of an "urban water cycle" as a microcosm of the bigger water cycles in nature.

When Orange County got into water reuse, the water district made sure key politicians were sold on the idea, and also reached out to local communities that tended to distrust the government -- mother's groups, Latinos, Vietnamese -- with the water district sending out engineers and using local doctors to help deliver the message. The water district is now planning an expansion of the Groundwater Replenishment System and going through the same drill. Officials see resistance crumbling, particularly among the younger generation.

No culture has any particular liking for human wastes, but some have a lower threshold of disgust than others. In China and other places in East Asia, human feces have long been used as fertilizer, and China is notably enthusiastic about biogas production from human and animal feces. India, in contrast, has a notably high threshold, with those who tend latrines having low social status. Western societies were once friendlier to the use of human wastes, also spreading human manures on lands, but with the introduction of modern sanitation such practices went on the fade. Now the fussiness is in retreat, as demonstrated by the ecologically-friendly toilets discussed earlier in this series. Mindsets still remain a major obstacle, but given the clear advantages of the recycling of human body wastes, such attitudes are very likely to change over time. [TO BE CONTINUED]



* GIMMICKS & GADGETS: As reported by AVIATION WEEK, aircraft are littered with little signs and logos, and applying them is troublesome: stencils are time-consuming, decals tend to peel off, and if a warning sign is damaged it has to be restored. Now Color Craft, a little company in Tukwila, Washington, near the Boeing plant in Everett, has developed an "O3 Aircraft Marking Technology" that combines the best of both worlds: paints are laid down on what looks like a decal, with the paint pattern transferred to an aircraft and then covered with sealant.

The images can be as elaborate as desired and last as long as painted images -- because they are painted images. They can be applied to aircraft that have been repainted even if the base paint layer isn't completely dry, since the image "breathes", allowing the paint job to be completed more quickly.

* Also as reported by AVIATION WEEK, the eruption of the Eyjafjallajokull volcano in Iceland in 2010 -- discussed here at the time -- caused considerable disruption of air traffic, volcanic ash posing a dangerous threat to aircraft engines. In response, low-cost air carrier EasyJet began a collaboration with European jetliner manufacturer Airbus and Nicarnica Aviation -- a spinoff of the Norwegian Institute For Air Research -- to build a sensor system to detect clouds of volcanic ash.

The "Airborne Volcanic Object Identifier & Detector (AVOID)" prototype consists of a pod mounted on the wingtips of a jetliner, each pod containing an uncooled infrared imager. The imagers can spot concentrations of ash, or ice crystals, from a range of about 100 kilometers (60 miles), giving pilots several minutes' warning. Software takes inputs from the sensor pods, as well as flight parameters and GPS location coordinates, to map out the cloud, with the data transmitted to air traffic control centers.

AVOID is in the final phase of testing. EasyJet has pushed the AVOID program since the 2010 eruption disrupted flights in the North Atlantic region for several days, at painful loss to air carriers; the investment in AVOID is minor in comparison. EasyJet officials believe that only about 100 jetliners, 20 of them from EasyJet, would need to be fitted with the AVOID pods to ensure minimal interruption in flight service in the event of another Iceland eruption.

* As reported by BLOOMBERG BUSINESSWEEK, in consequence to the horrific collapse of the Rana Plaza complex in Bangladesh on 24 April 2013 that killed more than 1,100 people, multinational companies have a heightened interest in monitoring conditions at subcontracting firms in developing nations. Currently, multinationals tend to use third-party audits, but subcontractors have been inclined to game the system, coaching employees on what to say and threatening to fire them if they don't play along.

Kohl Gill, a Silicon Valley entrepreneur, has a better idea. His firm LaborVoices provides services to allow factory workers to report on work conditions anonymously by phone, leaving voice messages in their native language that are tallied by the system, which also provides them with useful information. Multinational companies sign up for the service, which is priced much lower than the charges levied by third-party audit services, and get an online dashboard to track the data. Gill says that the data are "anonymized, aggregated across workers, and vetted for accuracy."

LaborVoices works with local trade unions and nonprofits to get the word out to workers; working through subcontractor management is counterproductive, since workers don't trust them. For the moment, LaborVoices has only a few clients, in the US and India, but is expanding to China and Bangladesh. LaborVoices has a competitor of sorts named Labor Link, established by nonprofit Good World Solutions of Oakland, California, which has surveyed tens of thousands of workers in Brazil, China, India, Mexico, Peru since going online three years ago, has clients including Marks & Spencer, Cisco Systems, and Patagonia.

Neither operation pays workers for information collected; it would inevitably bias the system. Officials of both firms say they are not attempting to replace trade unions and government regulatory organizations. All LaborVoices and Labor Link can do is provide information; somebody else must provide enforcement.



* PATENT SLEUTHING: Fussing over intellectual property (IP) is nothing new; as long as patent laws have been around, people have squabbled over them in court. Patent feuds have, however, become more of a nuisance in recent decades as "patent troll" companies obtain patents in bulk -- not to actually produce anything, but as a pretext to press patent suits against those who really are trying to produce something.

As reported by an article from BLOOMBERG BUSINESSWEEK ("Recruiting An Army To Battle The Patent Trolls" by Olga Kharif, 21 January 2013), firms under assault by patent trolls now have some help, in the form of Article One Partners, based in New York City. When Dutch electronics giant Philips was confronted with a potential patent challenge on its LED lighting products, the company turned to Article One for help. Working through its website, Article One asked for help from the tens of thousands of researchers that hang around there; their investigations quickly found a prior patent that shot down the patent claim.

According to Ruud Peters, the chief IP officer at Philips, thanks to Article One, Philips was able to "completely eradicate" the claim against it. Philips ended up paying Article One about $100,000 USD for the service, but it was a bargain at the price, Peters saying: "That's small money compared to the millions you might spend on litigation."

Article One does charge one-time fees, but also offers subscription services. The work is mostly done by volunteers -- engineers, academics, and retired patent attorneys -- out to pocket a little extra money, Article One handing out a few thousand dollars to each contributor who comes up with something valuable. IP experts say the crowdsourced model works very well; given its parallelism, it actually works better than an in-house staff of IP attorneys. Article One does have competitors such as Patexia and Ask Patents, but Article One is the biggest player, having conducted hundreds of studies and passed out millions of dollars to researchers. In 2012, Philips alone hired Article One to check into 33 patents, three times as many as in 2011.

CEO Cheryl Milone, a patent attorney, founded Article One in 2008, and has been able to attract substantial financial backing. The company expects to get a boost from the 2011 Leahy-Smith America Invents Act, which allows researchers and patent attorneys to electronically file evidence related to pending applications to the US Patent & Trademarks Office. Article One plans to make use of the America Invents Act and believes it should do much to put a cork into patent trolls.

While none of the researchers get rich off of their work for Article One, according to Milone about 8% make over $50,000 USD a year, which is pretty good pocket money. David Sarokin, a 60-year-old ecologist living in Washington DC, said he made almost $15,000 USD in six months, working a few hours a day. Sarokin added: "It certainly has been worth it for me. I have a bit of a Sherlock Holmes personality. I like to find information."



* VIRTUAL HUMANS FOR CRASH TESTING: Traditionally, crash testing for automobiles and other vehicles has relied on crash-test dummies, or "anthropomorphic test devices (ATD)" as they are formally known, described here in 2008. As discussed by an article from BBC WORLD Online ("Crash Tests: Not For Dummies" by Paul Rubens, 9 April 2013), ATDs are useful, but they have limitations -- although the popular Hybrid III ATD comes in a number of different configurations, it can't really model the full range of possible drivers and passengers, and not being flesh and blood, it doesn't honestly represent how flesh and blood reacts to a crash.

The ATD has still proven very useful, and it was the best thing available in its time -- but we have other options in the 21st century. The "Global Human Body Models Consortium (GHBMC)", a group of car manufacturers and suppliers, including Chrysler and General Motors, is working to come up with something better: "virtual humans", computer models of humans that are much more humanlike, "biofidelic", than ATDs. Scott Gayzick, a scientist at the Wake Forest Center for Injury Biomechanics in Virginia who is involved with the GHBMC, commented: "We want to be able to run tests with models of people who are obese, or with extra mass where real people have extra mass -- around the abdomen -- and we can't do that with an ATD."

To construct the models, GHBMC researchers begin by taking computed tomography (CT) scans of people in both lying and seated positions, to build up a 3D representation of a human skeleton; they then add magnetic resonance imaging (MRI) scans of people in a seated position using an upright MRI machine, to add the muscles, internal organs and brain. Joel Stitzel, one of the leaders of the GHBMC team working at Wake Forest, said: "Once we have the bone component and the soft-tissue component, we can build an accurate image of a person. That's like the canvas. Then we add the mechanics and the stiffnesses of the components to make it like a real human: we play Frankenstein."

So far, the GHBMC has completed a model of an average "50th percentile male"; despite its complexity, the model is only about a megabyte in size. The group is now working towards a more sophisticated model that can easily be scaled up or down and adjusted to simulate people of different sizes, builds, and ages. Once written, the virtual humans are placed in a computer model of a car under development, and then put through simulated accidents. Since cars are digitally designed these days, virtual crash testing is merely an extension of what is being done now.

A typical car accident occurs in a tenth of a second; a virtual simulation, in contrast, can take up to ten hours, running on a powerful 64-bit processor system. It takes so long because of the huge amounts of data involved. Analyzing a crash test with an ATD is simple from a data analysis point of view; a virtual collision provides a flood of data points to give a far more comprehensive picture of what happens. In addition, as as a virtual human model becomes more detailed, the more individual it becomes and less representative of a range of real-world humans. That means running virtual collisions with a range of variations of virtual humans.

However, it's still cheaper to perform virtual crashes than real ones. According to Guy Nusholtz, a manager at the Experimental and Computational Mechanics department at Chrysler who is interested in brain injuries: "When you run a test with a dummy, all you can say is that the dummy gets hit in the head, here are the forces, here is the acceleration and here is the motion-time history of the head. With a computer model, we can attempt to duplicate the fundamental physics of the brain and see what happens to it."

At least at the outset, regulatory agencies are necessarily going to be suspicious of computer models, and demand physical tests with ATDs to ensure that safety regulations are being met. GHBMC researchers shrug: virtual humans can be used in the product design phase to make sure the design is as safe as possible, with ATDs then used to validate the final product design. By ensuring that the product is properly designed towards safety, testing is simplified, reducing development costs. A safer product also means fewer liability suits.

The researchers see their work as having broader application, Gayzick saying: "We are excited about the possibility of computer games companies using our human models. They have their own models already, but their movements are not as realistic, and they are not internally accurate." Then again, since when are videogame heroes and heroines supposed to look like real people?



* ANOTHER MONTH: Early in May my DSL modem, which hooks me up to the internet, started going erratic, dropping my internet connection on occasion. It was worse Monday morning; I checked with the service provider, they didn't have a problem at their end.

Ah, I figured I had to replace the modem. Since the closest shop of the service provider was just north of Denver, I had to drop my normal work schedule and take the time to go south. Normally Tuesdays is my "day off" -- go to the supermarket, go to the library to restock on the latest magazines, and that week go to the bank to make a deposit. Since Monday was going to be shot for getting any real work done, I decided I'd run all my errands that day instead.

I had been thinking of taking a day off to go to the Colorado Springs zoo -- about 2.5 hours drive south of Loveland -- but though had I wavered on the notion, I knew I couldn't spare the time. As a consolation, when I went down to Denver I went to the zoo there. Yes, a very familiar place, but the nice thing about zoos is that, given a large enough zoo, something different is going on each time one visits.

Anyway, I came back home and installed the modem, with everything seeming okay. Actually, the problem was intermittent, and came back again in the morning; I contacted the service provider, and they sent out a tech the next day. He changed the port on the access multiplexer downstream and all was fine again. I put the old modem back; it's also a wi-fi hub, I didn't feel like reconfiguring my household wi-fi network, and I figured I should run the old modem to death anyway. I didn't mind having a second modem as a backup; I'm sure it has a good shelf life, and it's useful to have a spare for troubleshooting.

* This incident seemed to touch off a spate of malfunctions. In midmonth, I looked in the cabinet space under my kitchen sink on a Sunday afternoon and found it flooded with water; what gives? Turned out that my garbage disposal was leaking, presumably having lost a gasket. Aw nuts, swapping out a garbage disposal is a pain .... I went out and bought a new one on Tuesday; I had bought cheap for the one that failed, so this time I went to a higher-end item.

The installation was indeed a pain, but it was mostly because I hadn't recorded the difficulties the last time I did it. There are two main assemblies for a garbage disposal: a sink mounting fixture and the disposal itself. The sink mounting fixture consists of a drain that fits through the top of the sink basin, clamped in place by subassemblies on the bottom; the clamping assembly is held by a retaining ring snapped around the bottom of the drain, and then snugged up tight with three screws from below. There's a mounting ring on the bottom of the fixture, with a mating ring on the top of the disposal for attaching it to the fixture.

I got a bit confused in removing the old disposal, first unsnugging the mounting fixture, which was precisely the wrong thing to do. All that was necessary to remove the disposal was to rotate its mounting ring -- but which way to turn the ring? I finally figured out I could feel around the top of the mated rings for the "ramps" over which the disposal ring snugged up, and the slope confirmed which way to turn. Persistent if awkward application of a hammer got the disposal off -- after considerable expenditure of time, along with chipped fingernails, scratches, and ill humor. I later bought a little hammer for jobs like that in the future, my big claw hammer being inconvenient for relatively precise work in close quarters; I never pass up buying a tool if I vaguely think I need it.

Installing the new disposal was not such a problem, the only difficulty being installation of the mounting fixture -- I had to hold the clamping assembly together underneath the sink with one hand and snap in the retaining ring with the other. That was frustrating, until I remembered the miracle tool, duct tape. I taped the assembly to the bottom of the sink and then snapped in the retaining ring; at least I did one thing clever. Anyway, that done, attaching the disposal was straightforward, helped by a tool provided with the unit that fit into tabs on the ring on top of the disposal to turn it with. Having installed the disposal, I left the tool by the disposal; no point hunting around to find the thing the next time I replaced the disposal.

* Following that, I had some car problems, to be discussed next month. Incidentally, the trip to the Denver zoo was very worthwhile. I shot 173 photos there, threw out about a third right away, probably keep about 30. When I was at the sloth bear pit, somebody started throwing food in, which startled me -- people aren't suppose to do that, are they? Ah, it was a zookeeper. She then went to the grizzly bear pit next door and did the same. I asked her: "What's that, some sort of jerky?"

"That's right."

"Do they recognize individual keepers?"

"Yes, they're very bright."

On the way out, I passed the seal pool. A keeper was putting on a bit of a show with Lucy, one of the zoo's California sea lions. She was very well trained; the keeper was running her through a routine and one could hardly see the keeper giving her cues.

Lucy takes a dive

They have a harbor seal and the keeper was pointing out the difference between sea lions and seals. One I knew, sea lions have external ears, seals just have earholes. One I didn't quite realize is that sea lions get their propulsion from their forward flippers, which are much larger than the rear (Lucy lifts up a flipper to show it off) while seals get their propulsion from the rear flippers. Of course I knew that sea lions get around much easier on land than seals -- with Lucy humping and squirming forward, pretending to be a seal. I had the damnedest sense she thought it was great joke to make fun of seals. Incidentally, fur seals are actually a species of sea lion.

My timing on the zoo trip was perfect. I got there about 1:30 PM, left about 3:00 PM, traffic was starting to build but I never got gridlocked. If I'd left an hour later, I would have been trapped. Denver's not so bad -- bad is San Francisco, Seattle, and Washington DC. Trying to drive in DC is insane, partly because of the weird radial-overlaid-on-grid street layout; I'm told it was set up in that way to allow troops to move rapidly from one point to another and give them clear fields of fire. Next time I go there, I will NOT drive my car within the Beltway that rings the inner city.