jueves, 31 de mayo de 2018

All life on Earth, in one staggering chart

Scientists estimated the mass of all life. It’s mind boggling.
By weight, human beings are insignificant.

If everyone on the planet were to step on one side of a giant balance scale, and all the bacteria on Earth were to be placed on the other side, we’d shoot violently upward. That’s because all the bacteria on Earth combined are about 1,166 times more massive than all the humans.

Comparisons to other categories of life similarly demonstrate how very, very small we are. As a sweeping new study in the Proceedings of the National Academy of Sciences finds, in a census sorting all the life on Earth by weight (measured in gigatons of carbon, the signature element of life on Earth), we make up less than 1 percent of life.

There are an estimated 550 gigatons of carbon of life in the world. A gigaton is equal to a billion metric tons. A metric ton is 1,000 kilograms, or about 2,200 pounds.

We’re talking in huge, huge, mind-boggling terms here.

So, using the new data in PNAS, we tried to visualize the weight of all life on Earth to get a sense of the scale of it all.

All life on Earth, in one chart
What you’ll see below is a kind of tower of life. Each large block of this tower represents a gigaton of life, and the blocks are grouped into broad kingdoms. There are the protists (think microscopic life like amoebae), archaea (single-celled organisms somewhat similar to bacteria), fungi (mushrooms and other types of fungus), bacteria (you’re familiar with these, right?), plants, and animals.

As you can see, plants dominate our world. If the tower of life were an office building, plants would be the main tenants, taking up dozens of floors. Comparatively, all the animals in the world — seen in gray in the tower — are like a single retail shop (a trendy one, to be sure) on the ground floor.

The biomass distribution on Earth
Yinon M. Bar-On, Rob Phillips, and Ron Milo
And if we zoom in on all animal life, we again see how insignificant humans are compared to everyone else in the kingdom. Arthropods (insects) outweigh us by a factor of 17. Even the mollusks (think clams) weigh more.

What’s missing from this chart is just as important
Yet despite our small biomass among animals, we’ve had an overwhelmingly huge impact on the planet. The chart above represents a massive amount of life. But it doesn’t show what’s gone missing since the human population took off.

The authors of the PNAS article estimate that the mass of wild land mammals is seven times lower than it was before humans arrived (keep in mind it’s difficult to estimate the exact history of the number of animals on Earth). Similarly, marine mammals, including whales, are a fifth of the weight they used to be because we’ve hunted so many to near extinction.

And though plants are still the dominant form of life on Earth, the scientists suspect there used to be approximately twice as many of them — before humanity started clearing forests to make way for agriculture and our civilization.

The census in the PNAS paper isn’t perfect. Though remote sensing, satellites, and huge efforts to study the distribution of life in the ocean make it easier than ever to come up with estimates, the authors admit there’s still a lot of uncertainty. But we do need a baseline understanding of the distribution of life on Earth. Millions of acres of forests are still lost every year. Animals are going extinct 1,000 to 10,000 faster than you’d expect if no humans lived on Earth. Sixty percent of primate species, our closest relatives on the tree of life, are threatened with extinction.

We have to know how much more we stand to lose.

By Brian Resnick and Javier Zarracina
May 29, 2018

miércoles, 16 de mayo de 2018

Plug-and-play diagnostic devices

Jose Gomez-Marquez, co-director of MIT’s Little Devices Lab, holds a sheet of paper diagnostic blocks, which can be easily printed and then combined in various ways to create customized diagnostic devices. Image: Melanie Gonick/MIT

Modular blocks could enable labs around the world to cheaply and easily build their own diagnostics.

Researchers at MIT’s Little Devices Lab have developed a set of modular blocks that can be put together in different ways to produce diagnostic devices. These “plug-and-play” devices, which require little expertise to assemble, can test blood glucose levels in diabetic patients or detect viral infection, among other functions.

Ampli blocks. Image by Melanie Gonick, MIT. 
Our long-term motivation is to enable small, low-resources laboratories to generate their own libraries of plug-and-play diagnostics to treat their local patient populations independently,” says Anna Young, co-director of MIT’s Little Devices Lab, lecturer at the Institute for Medical Engineering and Science, and one of the lead authors of the paper.

Using this system, called Ampli blocks, the MIT team is working on devices to detect cancer, as well as Zika virus and other infectious diseases. The blocks are inexpensive, costing about 6 cents for four blocks, and they do not require refrigeration or special handling, making them appealing for use in the developing world.

We see these construction kits as a way of lowering the barriers to making medical technology,” says Jose Gomez-Marquez, co-director of the Little Devices Lab and the senior author of the paper.

Elizabeth Phillips ’13, a graduate student at Purdue University, is also a lead author of the paper, which appears in the journal Advanced Healthcare Materials on May 16. Other authors include Kimberly Hamad-Schifferli, an associate professor of engineering at the University of Massachusetts at Boston and a visiting scientist in MIT’s Department of Mechanical Engineering; Nikolas Albarran, a senior engineer in the Little Devices Lab; Jonah Butler, an MIT junior; and Kaira Lujan, a former visiting student in the Little Devices Lab.

Advanced Healthcare Materials Paper (click to enlarge)
Customized diagnostics

Over the past decade, many researchers have been working on small, portable diagnostic devices based on chemical reactions that occur on paper strips. Many of these tests make use of lateral flow technology, which is the same approach used in home pregnancy tests.

Despite these efforts, such tests have not been widely deployed. One obstacle, says Gomez-Marquez, is that many of these devices are not designed with large-scale manufacturability in mind. Another is that companies may not be interested in mass-producing a diagnostic for a disease that doesn’t affect a large number of people.

The Little Devices Lab researchers realized that they could get these diagnostics into the hands of many more people if they created a kit of modular components that can be put together to generate exactly what the user needs. To that end, they have created about 40 different building blocks that lab workers around the world could easily assemble on their own, just as people began assembling their own radios and other electronic devices from commercially available electronic “breadboards” in the 1970s.

When the electronic breadboard came out, that meant people didn’t have to worry about building their own resistors or capacitors. They could worry about what they actually wanted to use electronics for, which is to make the entire circuit,” Gomez-Marquez says.

In this case, the components consist of a sheet of paper or glass fiber sandwiched between a plastic or metal block and a glass cover. The blocks, which are about half an inch on each edge, can snap together along any edge.
  • Some of the blocks contain channels for samples to flow straight through, 
  • some have turns, and 
  • some can receive a sample from a pipette or 
  • mix multiple reagents together.
The blocks can also perform different biochemical functions. Many contain antibodies that can detect a specific molecule in a blood or urine sample. Those antibodies are attached to nanoparticles that change color when the target molecule is present, indicating a positive result.

These blocks can be aligned in different ways, allowing the user to create diagnostics based on one reaction or a series of reactions. In one example, the researchers combined blocks that detect three different molecules to create a test for isonicotinic acid, which can reveal whether tuberculosis patients are taking their medication.

The blocks are color-coded by function, making it easier to assemble predesigned devices using instructions that the researchers plan to put online. They also hope that users will develop and contribute their own specifications to the online guide.

Better performance
The researchers also showed that in some ways, these blocks can outperform previous versions of paper diagnostic devices. For example, they found that they could run a sample back and forth over a test strip multiple times, enhancing the signal. This could make it easier to get reliable results from urine and saliva samples, which are usually more dilute than blood samples, but are easier to obtain from patients.

These are things that cannot be done with standard lateral flow tests, because those are not modular — you only get to run those once,” says Hamad-Schifferli.

The team is now working on tests for human papilloma virus, malaria, and Lyme disease, among others. They are also working on blocks that can synthesize useful compounds, including drugs, as well as blocks that incorporate electrical components such as LEDs.

The ultimate goal is to get the technology into the hands of small labs in both industrialized and developing countries, so they can create their own diagnostics. The MIT team has already sent them to labs in Chile and Nicaragua, where they have been used to develop devices to monitor patient adherence to TB treatment and to test for a genetic variant that makes malaria more difficult to treat.

Catherine Klapperich, associate dean for research and an associate professor of biomedical engineering at Boston University, says the MIT team’s work will help to make the diagnostic design process more inclusive.

By reducing the barriers to designing new point-of-care paperfluidics, the work invites nonexperts in and will certainly result in new ideas and collaborations in settings all around the world,” says Klapperich, who was not involved in the research. “The practical demonstrations of the system presented here are poised to be immediately useful, while the possibilities for others to build on the tool are large.

The researchers are now investigating large-scale manufacturing techniques, and they hope to launch a company to manufacture and distribute the kits around the world.

We are excited to open the platform to other researchers so they can use the blocks and generate their own reactions,” Young says.

The research was funded by a gift from Autodesk and the U.S. Public Health Service.

By Anne Trafton | MIT News Office 
May 16, 2018

This Hard-to-Destroy Drone Goes From Rigid to Flexible When It Crashes

Image: EPFL Don't worry, the drone is fine!
Anyone who’s ever flown a drone of any sort will tell you that sooner or later, you’re going to crash it. The question is how exactly you will go about doing this, and how much of the drone will be functional after it’s happened. Most flying animals somewhat frustratingly don’t have this problem: Birds and insects run into things occasionally (or all the time, for small bugs), and just shrug it off and keep on going, thanks to their biological design, which includes both stiffness and flexibility. Now roboticists at the EPFL, in Lausanne, Switzerland, are relying on these same qualities to design a highly resilient quadrotor that’s impressively difficult to destroy.

There are three primary strategies for designing drones with impact resistance.
  1. The first is to just protect the propellers by surrounding them with the frame of the drone or with individual propeller guards. Most commercial drones have something like this. With this level of protection, you’re less likely to injure people, but since the prop protection is rigid, you’re more likely to injure the drone itself if (I mean, when) you crash it. The EPFL quadrotor uses a flexible frame that locks in place with magnets. When a collision occurs, the frame breaks away from the magnets, and once the energy is dissipated, elastic bands pull the frame back together and you’re good to go.
  2. The second level of impact protection is to design your drone in a way that it can absorb energy from the crash without breaking into pieces. One way of doing that is to decouple the frame of the drone by, say, using flexible, elastic couplers. This gives you a “squishy” drone, which is very effective at handling impacts, but it’s also squishy in-flight, which causes all kinds of structural and stability problems
  3. The most impressive level of impact protection that we’ve seen in drones is the brute force approach of just surrounding the entire thing with a flexible, rotating cage. Flyability has made a compelling business case for using drones with protective cages, and for some applications, it’s fantastic. You do, however, pay a penalty, since the cage can increase the overall size of the drone by upwards of 60 percent, meaning that it’s safer to run into things, but you’re also much more likely to run into things. The cage adds mass as well, leading to a drone that can’t lift as much or fly as far.

EPFL’s idea is a compromise of sorts: They’ve managed to create a drone with a frame that’s rigid right up until it smashes into something, at which point it turns flexible:

The inspiration for this design came from insect wings. To fly, insects need wings with high stiffness, but flexibility is critical for absorbing shock, and wasps do it with a special joint that allows the entire wing to “reversibly crumple” during a collision, as the EPFL researchers explain:

Wasp wings display dual stiffness, that is the ability to reversibly transition between rigid and soft states, which provides mechanical resilience without impairing flight performances. The wings of wasps contain a flexible resilin joint... This design allows the wing tip to slightly flex during flight (rigid state), but reversibly crumple… during collisions (soft state). If the dual-stiffness behavior is impaired... the rigid wings undergo severe tear during collision. Therefore, this design provides crash resilience by effectively preventing wing overload during collisions without compromising flight capabilities.
Image: EPFL The quadcopter has two main parts: the external frame and a central case. They are held together by magnetic joints, each with two magnets and a spring (inset).
The quadrotor that the researchers came up with uses a flexible frame that locks in place around the core of the quadrotor with magnets. When a collision occurs, the frame breaks away from the magnets, absorbing the energy of the collision. Once the energy is dissipated, elastic bands pull the frame back into its original configuration, and the magnets snap together again, and you’re good to go.

The uniqueness of the proposed design lies in the fact that the frame is rigid during flight, but softens during collisions. This allows combination of the advantages of both rigid and soft systems: stability and rapid response to user commands during flight, leading to flight performance equivalent to a drone equipped with a standard rigid frame, and crash resilience like a soft system. The experiments showed a satisfying survivability of the frame of the drone, that withstood roughly 50 collisions with no permanent damage.

The researchers, who are members of EPFL’s Laboratory of Intelligent Systems, suggest that an approach like this could be useful for all kinds of robots, not just drones. Most robots are rigid for various performance reasons (like precision), but the ability to be flexible when necessary without using intrinsically soft materials could could come in handy for grasping, locomotion, or simply as an added safety measure for human-robot interaction tasks.

S. Mintchev, Member, IEEE, S. de Rivaz and D. Floreano, Senior Member, IEEE
Insect-Inspired Mechanical Resilience for Multicopters,” by S. Mintchev, S. de Rivas, and D. Floreano from EPFL was published in the 25 January 2017 issue of IEEE Robotics and Automation Letters.

[ EPFL ]

By Evan Ackerman
Posted 9 Mar 2017