Almost a year ago, I tore the folks at Lawrence Livermore National Lab a new one over their security policies for a computer algorithm called NEC-4. The short version is that this is a very useful antenna modeling algorithm developed with government funds, but LLNL keeps it locked behind a seemingly absurd paywall. Students, amateur scientists, and would-be entrepreneurs interested in wireless technology have to pay a steep entry fee if they want access to this highly useful tool, even though it was developed entirely with taxpayers’ money.
Recently, though, I received a note from someone at another research center who explained the situation much better than the LLNL folks do. His name is Jim. Here are some excerpts from our exchange:
I work at JPL, and it costs us substantially more than $300 to distribute export controlled software, just to handle the paperwork. There’s at least 3 people involved: the “software release authority” who deals with all software distribution; the export control record keeping person; and me. One of the first two runs the name/business name through some search databases. Pretty fast, but by the time you’re done, you’ve probably consumed a couple or 3 work hours.
Part of the problem is that even for non-export controlled software, there’s a record keeping requirement imposed by Congress. They want to know how many people are benefiting from using the government developed software so we have to keep records so we can report to some nameless entity who can summarize it in a report that probably never gets read except when someone complains to the IG or when some Congressperson gets interested.
Apparently I was mistaken about the actual costs of distributing this software. I stand corrected: it probably does really cost LLNL a few hundred dollars to provide a copy, solely because the algorithm is covered by US export control laws. More on that in a moment.
Jim also commented on my contention that the “security check” required to obtain the algorithm is meaningless. It seems that I was more or less correct about that:
As for “paying a U.S. person to get the CDROM and forward it,” that’s a pretty clear violation of the export control laws. The person doing this is setting themselves up for pretty severe penalties if the govt got its dander up. Oddly, not much is required in terms of authenticating or verifying that the person you give it to is actually a U.S. Person under the law. Their statement that they are is sufficient. They can lie, and that puts the violation on the recipient, not the sender.
Last I checked, the countries and organizations we need to worry about are universally willing to lie and break our laws, so “put[ting] the violation on the recipient” accomplishes nothing.
In other words, export control laws mean that LLNL has to keep this useful algorithm locked behind a paywall, and has to cite “security” as a justification for that, but it would be much easier for everyone – and no less secure – to make the thing freely available.
There may be a way out of this silliness, though:
In reality, what someone should do is see if they can reclassify it as not export controlled. If you were to seek a determination for NEC-4 today, I doubt it would wind up controlled. It’s unclear to me what parts of it triggered the controls in the first place, but probably it’s the parts with improved models for insulated antennas submerged in lossy dielectric mediums (like seawater) and that probably triggered the export control rating. However, I would think that stuff has been published in the open literature by now.
It sounds like a properly motivated member of Congress could accomplish a lot of good here. Removing NEC-4 from export control and making it freely available would allow amateur experimenters, students, and other interested folks to learn more about antenna modeling and maybe come up with some new ideas. It would also enable the Open Source community to work on the underlying algorithm, possibly improving it. With breakthroughs in wireless communication now producing massive economic benefits, these are topics that could clearly use more brains. I certainly plan to make that argument in a letter to my representatives. If you agree, here’s where you can find your Congressional representative, and here’s the contact information for your Senator.
Last Monday, the night before I was due to cover a big conference, my nearly brand-new MacBook Air died: suddenly, completely, and unrecoverably. After hurling a stream of expletives in the general direction of Cupertino, I paid a visit to the Apple store conveniently located next door to the conference hotel. They confirmed that my MacBrick was really most sincerely dead, but there was no way they’d be able to fix it before the end of the conference.
Before leaving the store, though, I picked up a Bluetooth keyboard. That, plus my iPhone and a few other items I already had with me, became my new laptop. I was amazed at how well the arrangement worked, and also at how many people came over to comment on my setup. Here’s what it looked like:
Considering its $110 price tag, it’s a surprisingly capable system. The Pages app stores files on the iCloud service, so they’re automatically backed up and accessible from anywhere. The keyboard is very quiet and has a feel I’m accustomed to, and the tripod adjusts to hold the phone at whatever angle I need. Of course it also has a built-in web browser, email client, and the ability to install pretty much any other kind of program one might need.
I’m not sure I’d want to rely on this as my only computer on a business trip, but I’ll certainly pack the components for it in case I need a backup again.
Oh, and my regular laptop is once again among the living – my local Apple store fixed it under warranty after I got home.
Last weekend’s nor’easter utterly destroyed the electrical infrastructure of Western Massachusetts, where I live. Please do not expect prompt comment approvals or email replies until next week, at the earliest. Thanks.
“Microfluidics” is one of the hottest buzzwords in biotechnology and diagnostics research these days, with good reason: these lab-on-a-chip devices are about the coolest technology to come along since monoclonal antibodies. The designs vary widely, but the basic principle is to take traditional lab assays and miniaturize them onto silicone or plastic chips, often using manufacturing techniques developed by the semiconductor industry. I’ve blogged about these nifty devices once or twice (okay, maybe three times) before.
While I’ve found the technology fascinating to watch, it’s remained a bit of a laboratory curiosity. Everyone seems to agree that the “killer app” for microfluidics will be field-portable devices that that will let minimally-trained people diagnose diseases or detect specific compounds in the environment, especially in poor countries. That’s because chip-based labs can incorporate all of their equipment and reagents onto a disposable device no larger than a credit card. In principle, a technician could place a drop of fluid, such as blood, onto one end of the chip, and micrometer-size channels would siphon it around, mixing it and moving it to different chambers to perform assays that would normally require a fully-equipped lab. The volumes are so small that the reactions tend to occur very quickly, often shortening multi-hour tests to a few minutes. But that’s where the good news ends.
Because the reactions produce subtle chemical changes inside minuscule containers, detecting the result usually requires sophisticated analytical equipment, at which point we’re right back to building a full-size laboratory. No matter how cheap or portable the chips get, the assay readout has generally remained huge and pricey.
Until now. Two recent papers highlight what I think is a new stage in the development of microfluidics, where researchers are finally addressing the readout problem. In one effort, scientists at Columbia University report on a microfluidic clinical testing system that incorporates a whole slew of new ideas. More importantly, it actually seems to work in the field. As senior investigator Samuel Sia says in an accompanying press release:
“We have engineered a disposable credit card-sized device that can produce blood-based diagnostic results in minutes,” said Sia. “The idea is to make a large class of diagnostic tests accessible to patients in any setting in the world, rather than forcing them to go to a clinic to draw blood and then wait days for their results.”
Sia’s lab at Columbia Engineering has developed the mChip devices in collaboration with Claros Diagnostics Inc., a venture capital-backed startup that Sia co-founded in 2004. The microchip inside the device is formed through injection molding and holds miniature forms of test tubes and chemicals; the cost of the chip is about $1 and the entire instrument about $100.
The injection-molding process is a departure from most microfluidic construction methods. Instead of engraving the device onto a silicon chip, the researchers made a mold and cast duplicates in a mass-production system. Injection molding gives us plastic cups and soda bottles, so it’s clearly a mature technology that can be scaled way, way up. That’s what drives the per-chip cost down so low.
The investigators used this cheap chip to build a miniaturized ELISA, or enzyme-linked immunosorbent assay, platform. If you’ve ever been tested for any infectious disease, you’ve probably had an ELISA; it’s one of the most common and important assays in clinical diagnosis. To perform it, one needs to incubate a sample with a reagent that will bind some analyte – let’s say an antigen that will bind antibodies against HIV in a patient’s blood. Once the analyte binds, a series of washes and secondary reagents clears up background reactions and causes some kind of easily-detected chemical change. It typically takes a skilled lab technician a few hours to perform an ELISA, and it requires careful attention to detail through the various washing and incubation steps.
On the new chips, a simple channel meanders through the plastic. The binding reagent is stuck to one section of the channel, and Sia and his colleauges feed the blood sample, wash solutions, and other reagents through the tube sequentially. To separate the reagents, they simply added tiny bubbles between them, like you might see in a very small straw that’s reached the bottom of the glass. A common medical syringe provides the vacuum force to draw the whole train of reagents through the system.
Completing this tour de force of clever ideas, the team used a nanoparticle-based detection system that deposits visible quantities of silver in the channel if there’s been a reaction. A cheap absorbance meter quantifies the amount of silver, and determines whether a test is positive or negative. The researchers walk through the system’s advantages in a nicely-produced video interview Nature Medicine released to accompany the piece:
As you’ll see in the video, the researchers also put their system to the ultimate test, hauling it to Rwanda and testing actual patient samples in an underfunded, overworked clinic. The results were impressive: the new assay is about as accurate as traditional ELISA tests for detecting HIV and syphilis, but much faster and cheaper.
This makes me wonder whether we’re about to see another example of leapfrogging in poor countries. The most popular (and really only) current example of this phenomenon is cellular phones. There are virtually no landline connections in most poor countries, but nearly everyone has a phone. By missing the first telecommunication revolution, these countries have “leapfrogged” to the second, gaining all of the advantages of instant communication without going through the intermediate stages of rural electrification, Ma Bell, party lines, and rotary dials. If microfluidic devices can bring modern medical tests to the bedside in Rwanda, will we see them and other poor countries catapulting into 21st century medicine without having to establish 20th (or even 19th) century medical infrastructure first? Maybe.
What makes me optimistic about this is that Sia and his colleagues aren’t the only ones working on this problem. Indeed, around the time their paper came out, a less-noticed but equally interesting bit of work came out in the journal Analytical Chemistry. In that paper, Aydogan Ozcan and his colleagues at UCLA and elsewhere describe a system for performing flow cytometry on a microfluidic device. Flow cytometry, or cell sorting, is a sort of ELISA on speed. Rather than incubate the bulk sample with the reagent, cell sorters separate individual cells into a stream of droplets, like one would get by shaking a running garden hose. The droplets pass through a detector that measures specific parameters of the cell, such as its light diffraction characteristics or whether it bound a fluorescently labeled antibody. Researchers can then quantify exactly how many cells of each type were in a sample.
It’s a tremendously powerful technique for immunological research, and can also be used to perform a variety of blood-counting assays, but it requires even more skill and money than an ELISA. Research-grade cell sorters are massive machines that usually occupy a small room of their own and employ a dedicated technician.
Ozcan’s team decided to use a cell phone instead. With about $5 worth of parts, they cobbled together an adapter that connects an inexpensive microfluidic cell sorter to the camera on a Sony-Ericsson phone. As they explain in an accompanying press release:
The microfluidic assembly is placed just above a separate, inexpensive lens that is put in contact with the cell phone’s existing camera unit. This way, the entire cross-section of the microfluidic device can be mapped onto the phone’s CMOS sensor-chip. The sample fluid is delivered continuously through a disposable microfluidic channel via a syringe pump.
The device is illuminated from the side by the LEDs using a simple butt-coupling technique. The excitation light is then guided within the cross-section of the device, uniformly exciting the specimens in the imaging fluid. The optofluidic pumping scheme also allows for the use of an inexpensive plastic absorption filter to create the dark-field background needed for fluorescent imaging. In addition, video post-processing and contour-detection and tracking algorithms are used to count and label the cells or particles passing through the microfluidic chip.
While they haven’t taken it into a poor country’s clinics yet, the investigators did put the system through its paces in the lab. So far, they’ve demonstrated that it can measure white blood cell density as a cell sorter, and also operate as a mid-power fluorescent microscope. The former capability could provide tests for leukemia and AIDS progression, while the latter could be useful for a variety of analyses, including detecting pathogens in drinking water.
It’s going to take more than a couple of new testing systems to fix the health problems of poor countries, but papers like these – and I suspect others will follow shortly – show at least part of the solution. Perhaps these technologies will even make their way back to the developed world, as we seem to have some of our own issues with medical costs these days.
Chin, C., Laksanasopin, T., Cheung, Y., Steinmiller, D., Linder, V., Parsa, H., Wang, J., Moore, H., Rouse, R., Umviligihozo, G., Karita, E., Mwambarangwe, L., Braunstein, S., van de Wijgert, J., Sahabo, R., Justman, J., El-Sadr, W., & Sia, S. (2011). Microfluidics-based diagnostics of infectious diseases in the developing world Nature Medicine DOI: 10.1038/nm.2408
Seo, S., Isikman, S., Sencan, I., Mudanyali, O., Su, T., Bishara, W., Erlinger, A., & Ozcan, A. (2010). High-Throughput Lens-Free Blood Analysis on a Chip Analytical Chemistry, 82 (11), 4621-4627 DOI: 10.1021/ac1007915
With my old iPod Photo losing the ability to hold a charge, I finally got one of the newfangled versions. Searching for science-related items in the App Store, I was heartened to discover the Skeptical Science app. It does the same thing as the website of the same name. As the site’s creators explain:
Scientific skepticism is healthy. Scientists should always challenge themselves to expand their knowledge and improve their understanding. Yet this isn’t what happens in global warming skepticism. Skeptics vigorously criticise any evidence that supports man-made global warming and yet uncritically embrace any argument, op-ed piece, blog or study that refutes global warming.
So this website gets skeptical about global warming skepticism. Do their arguments have any scientific basis? What does the peer reviewed scientific literature say?
While the developers have included tons of current citations and links to reams of data, the app (and the web site) manage to organize all of this information in a very intuitive way. Just click on an argument and you get graphs, maps, and understandable explanations.
Unsurprisingly, denialists have already declared that the app is just another part of the conspiracy.
After an unfortunate incident involving a broken beer glass, the Landsdowne Pub in Boston has switched to plastic cups. Fortunately, I don’t frequent that establishment, but it seems the Commonwealth has forced a number of other pubs to use plastic containers as well. Blech. New Englanders should instead follow the lead of Old England on this issue: shatterproof pint glasses.
There’s a bill in the Massachusetts legislature right now that I really hope passes: it’s called the Right to Repair act. It would require automakers to publish all of the diagnostic codes and repair data for their vehicles, and make them available to anyone for a reasonable price. Independent auto shops and parts dealers are lobbying hard for it, and so are occasional tinkerers like me.
If you live in Massachusetts and are sick of having to take your car to the overpriced, often incompetent mechanics at the dealership, please write to your state Representative and tell them to pass this. If you’ve forgetten your Representative’s name, check the House site.
For folks in other parts of the country, feel free to cheer for us. Once again, Massachusetts is on the verge of doing something the rest of you will thank us for later.
A few days ago, Facebook executives responded to questions from New York Times readers. I just saw the article, and was struck by this quote from Elliot Schrage, the company’s vice president for public policy. He was answering (or at least purporting to answer) a question about why the site constantly adds absurd new features and turns them “on” for all users by default:
We know that changing Facebook — something people have demonstrated is important to them — can be unsettling. But we’re always trying to be better and do more for our users. Clearly, we need to rethink the tempo of change and how we communicate it. Trust me.
Trust me? To paraphrase Inigo Montoya, Facebook keeps using that phrase, but I do not think it means what they think it means.
In related news, I see that I’ve started a trend with my recent departure from the service. Apparently, Google has noticed that more people are now interested in deleting their Facebook accounts than in downloading YouTube videos.
Lab-on-a-chip devices are a hot topic these days, and more and more researchers are finding cheap ways to make these tiny devices (see my earlier posts on these cool tools for background). Now, some Australian scientists have taken the process into the arts and crafts department:
This paper describes a new and simple concept for fabricating low-cost, low-volume, easy-to-use microfluidic devices using threads. A thread can transport liquid via capillary wicking without the need of a barrier; as it is stainable, it is also a desirable material for displaying colorimetric results. When used in sewing, threads have 3D passageways in sewed materials. The wicking property and flexibility of thread make it particularly suitable to fabricate 3D microfluidic devices. Threads can also be used with other materials (e.g., paper) to make microfluidic devices for rapid qualitative or semiquantitative analysis. These thread-based and thread-paper-based devices have potential applications in human health diagnostics, environmental monitoring, and food safety analysis, and are particularly appropriate for the developing world or remote areas, because of their relatively low fabrication costs.
They’re not kidding about the low fabrication costs. In the paper, they stitched some simple “devices” out of ordinary cotton thread and paper. They did “de-wax” the thread in a vacuum plasma reactor, which most folks probably don’t have sitting around, but there could be other ways of accomplishing that step. The other limitation is that the threads average 244µm in diameter, which is a good bit fatter than the features in a typical commercial lab chip. Maybe we could call these “millifluidic” devices instead of “microfluidic.”
Still, it’s a very cool idea, and it could have some unusual applications. Now if you’ll excuse me, I have to go stitch up a salmonella-detecting dish cloth.