Category Archives: Science

2017 Gordon Bell Prize

In an earlier post, I noted the ascendance of China in High Performance Computing.

While the Top500 list can be dismissed as empty marketing, China’s rise is very real.

This year, for the second year in a row, the ACM Gordon Bell Prize for innovation was awarded to a Chinese team. This year’s honor recognizes  a new earthquake simulation.

Nice work, all!

The important point is that this prize represents technical prowess in making practical use of large scale computing. You need big iron to do it, but you get a Gordon Bell prize for having serious programming chops.

As I said, the Top500 is just the peak of the mountain. There is a lot of mountain all the way down.

China is not just building gigantic systems that are tuned to kill on the benchmarks, they are building a generation of brilliant scientists and engineers who are surpassing the rest of the world.

Well done!

  1. Haohuan Fu, Conghui He, Bingwei Chen, Zekun Yin, Zhenguo Zhang, Wenqiang Zhang, Tingjian Zhang, Wei Xue, Weiguo Liu, Wanwang Yin, Guangwen Yang, and Xioafei Chen, “18.9-Pflops Nonlinear Earthquake Simulation on Sunway TaihuLight: Enabling Depiction of 18-Hz and 8-Meter Scenarios.”. 2017, ACM Gordon Bell Prize.

Antarctica Heat Flux Map

One of the most important scientific questions of the early twenty first century is, “what’s going on in Antarctica?”

Antarctica is a the largest reserve of ice on the planet, and when (not if) the ice melts, it will raise sea levels by tens of meters. Glub.  (See a new NASA simulation of the effects of the melting ice.)

Just how fast is the ice melting?

This is a complex question to answer. The ice caps are gigantic (miles deep at places), and warmed by the air above and the Earth and sea underneath. Warmer air and water melt the ice, but may produce more new snow. There are liquid rivers and lakes under the ice which erode and melt from underneath. In some places glacier of ice are flowing down to the sea, where they will break up and melt.

It’s complicated.

This week a team of British researchers published a map that reflects an important piece of the picture: the heat flux under the ice [3]. This is the heat coming from the Earth’s interior, which they show is quite variable across the continent.

Hotspots are located under West Antarctica; in contrast, the East is broadly relatively cold. British Antarctic Survey.

The study used several measures of the magnetic properties of the rock under the Antarctic ice, including surface, air craft, and satellite surveys. Molten rock loses its magnetic field at a specific temperature, so the magnetic measurements can show where the rock cools below this limit. This can be used to infer the temperature at various depths below the surface.

The resulting map shows considerable variation across the continent. The warmest locations will presumably tend to melt more than cooler places (on the underside of the ice).

One interesting point from the map is that West Antarctica is melting faster than other areas, but the heat flux from the Earth is low. This suggests that the melting is due to warmer seas and ice flows, with little contribution from geothermal heat.

This dataset will contribute to many studies of the Antarctic ice. (It will be literally the foundation for many simulations.)

  1. Jonathan Amos, Antarctica’s warm underbelly revealed, in BBC News – Science & Environment. 2017.
  2. Eric Larour, Erik R. Ivins, and Surendra Adhikari, Should coastal planners have concern over where land ice is melting? Science Advances, 3 (11) 2017.

Yasmina M. Martos, Manuel Catalan, Tom A. Jordan, Alexander Golynsky, Dmitry Golynsky, Graeme Eagles, and David G. Vaughan, Heat flux distribution of Antarctica unveiled. Geophysical Research Letters:n/a-n/a,


Biomimetic Robotic Zebrafish

Bioinspired and Biomimetic systems are the bees knees (sometimes, literally! [1]).

In some cases, taking bio inspiration leads to designs and design principles for human purposes (e.g., crawly robots inspired by Earthworms [2], or nets inspired by spiderwebs [4]).

Other times, creating a biomimetic robot teaches us about nature.

A group of European researchers from Ecole Polytechnique Fédéral de Lausanne and Sorbonne report this fall on a project that has created a robot zebrafish (Danio rerio) that joins the school of live zebrafish [3].

This is actually pretty difficult, because zebrafish are kind of loosey-goosey about schooling, coming together as needed in different situations. Today’s successful zebrafish must pay attention to the other fish, and play nicely with others.

The result is a robot not only looks and swims like a zebrafish, it learns the social signals of the fish, and behaves correctly I.e., it mimics the anatomy, the movement, the behavior, and the social signaling of the natural fish.


This seemingly rather simple result required analysis of how zebrafish school. The researchers developed a two level model, a high level strategy (where the school is going) and a more detailed movement model (how to move in the school).

They also had to quantify the “social integration” achieved by the robot and other fish, which is a measure of how zebrafish-like the robot is, compared to observations of the real zebrafish.

And, of course, they used a fishbot that looks and swims like a zebrafish. For some reason, zebrafish aren’t fooled by a lure that is a very abstract fish shape.

The researchers emphasize that all three forms of mimicry are important for successful schooling.  She’s gotta look like a zebrafish, swim like a zebrafish, and follow along like a zebrafish.

These results suggest that it should be possible to create robots that not only join in, but persuade and lead a school via the natural signaling of the fish. Such a robot or group of robots presumably would be a low-stress method to herd fish. (I’m not completely sure why one would need to herd zebrafish, per se.)

This study is pretty awesome.

It does to seem like kind of a one-off case, though. It took a lot of work to observe and model these small groups of zebrafish. It isn’t clear how well these techniques might apply to larger groups, longer time periods, other environments, or other species.

Obviously, it will be useful to automate the learning of the social signals and so on as they suggest. Eventually, this might lead to a theory of fish—metaknowledge of different cognitive models in fish. Now that would be cool.

  1. J. Amador Guillermo, Matherne Marguerite, Waller D’Andre, Mathews Megha, N. Gorb Stanislav, and L. Hu David, Honey bee hairs and pollenkitt are essential for pollen capture and removal. Bioinspiration & Biomimetics, 12 (2):026015, 2017.
  2. Fang Hongbin, Zhang Yetong, and K. W. Wang, Origami-based earthworm-like locomotion robots. Bioinspiration & Biomimetics, 12 (6):065003, 2017.
  3. Leo Cazenille, Bertrand Collignon, Yohann Chemtob, Frank Bonnet, Alexey Gribovskiy, Francesco Mondada, Nicolas Bredeche, and José Halloy, How mimetic should a robotic fish be to socially integrate into zebrafish groups ? (accepted). Bioinspiration & Biomimetics, 2017
  4. Zheng, L., M. Behrooz, and F. Gordaninejad, A bioinspired adaptive spider web. Bioinspiration & Biomimetics, 12 (1):016012, 2017.



Robot Wednesday


PS. Wouldn’t  “Biomimetic Robotic Zebrafish” be a good name for a band?

Listening for Mosquitos

The ubiquitous mobile phone has opened many possibilities for citizen science. With most citizens equipped with a phone, and many with small supercomputers in the purse or pocket, it is easier than ever to collect data from wherever humans may be.

These devices are increasing the range of field studies, enabling the identification of plants and animals by sight and sound.

One key, of course, is the microphones and cameras. Sold to be used for deals and dating, not to mention selfies, these instruments are outstripping what scientists can afford.

The other key is that mobile devices are connected to the Internet, so data uploads are trivial. This technology is sold for commerce and dating and for sharing selfies, but it is perfect for collecting time and location stamped data.

In short, the vanity of youngsters has funded infrastructure that is better than scientists have ever built. Sigh.


This fall the Stanford citizen science folks are talking about yet another crowd sourced data collection: an project that identifies mosquitos by their buzz.

According to the information, Abuzz works on most phones, including older flip phones (AKA, non-smart phones).

It took me a while to figure out that Abuzz isn’t an app at all. It is a manual process. Old style.

You use the digital recording feature on your phone to record a mosquito. Then you upload that file to their web site. This seems to be a manual process, and I guess that we’re supposed to know how to save and upload sound files.

The uploaded files are analyzed to identify the species of mosquito. There are thousands of species, but the training data emphasized the important, disease bearing species we are most interested in knowing about.

A recent paper reports the details of the analysis techniques [2]. First of all, mobile phone microphones pick up mosquito sounds just fine. As we all know, the whiny buzz of those varmints is right their in human hearing, so its logical that telephones tuned ot human speech would hear mosquitos just fine.

The research indicates that the microphone is good in a range of up to 100mm. This is pretty much what you would expect for a hand held phone. So, you are going to have to hold the phone up to the mosquito, just like you would pass it to a friend to say hello.

At the crux of the matter, they were able to distinguish different mosquitos from recordings made by phone. Different species of mosquito have distinct sounds from their wing beats, and the research showed that they can detect the differences from these recordings.

They also use the time and location metadata to help identify the species. For example, the geographic region narrows down the species that are likely to be encountered.

The overall result is that it should be possible to get information about mosquito distributions from cell phone recordings provided by anyone who participates. This may contribute to preventing disease, or at least alerting the public to the current risks.

This project is pretty conservative, which is an advantage and a disadvantage. The low tech data collection is great, especially since the most interesting targets for surveillance are likely to be out in the bush, where the latest iPhones will be thin on the ground.

On the other hand, the lack of an app or a plug in to popular social platforms means that the citizen scientists have to invest more work, and get less instant gratification. This may reduce participation. Obviously, it would be possible to make a simple app, so that those with smart phones have an even simpler way to capture and upload data.

Anyway, it is clear that the researchers understand this issue. The web site is mostly instructions and video tutorials, featuring encouraging invitations from nice scientists. (OK, I thought the comment that “I would love to see is people really thinking hard about the biology of these complex animals” was a bit much.

I haven’t actually tried to submit data yet. (It’s winter here, the skeeters are gone until spring). I’m not really sure what kind of feedback you get. It would be really cool to return email a rapid report (i.e., within 24 hours). It should say the initial identification from your data (or possibly ‘there were problems, we’ll have to look at it), along with overall statistics to put your data in context (e.g., we’re getting a lot of reports of Aegyptus in your part of Africa).

To do this, you’d need to automate the data analysis, which would be a lot of work, but certainly is doable.

I’ll note that this particular data collection is something that cannot be done by UAVs. Drones are, well, too droney. Even if you could chase mosquitos, it would be difficult to record them over the darn propellers. (I won’t say impossible—sound processing can do amazing things).

I’ll also note that this research method wins points for being non-invasive. No mosquitos were harmed in this experiment. (Well, they were probably swatted, but the experiment itself was harmless.) This is actually important, because you don’t want mosquitos to selectively adapt to evade the surveillance.

  1. Taylor Kubota, Stanford researchers seek citizen scientists to contribute to worldwide mosquito tracking, in Stanford – News. 2017.
  2. Haripriya Mukundarajan, Felix Jan Hein Hol, Erica Araceli Castillo, and Cooper Newby Using mobile phones as acoustic sensors for high-throughput mosquito surveillance. eLife. doi: 10.7554/eLife.27854 October 11 2017,

The Psychology of the Supernatural

This month Kathryn Schulz writes a lovely little essay in the New Yorker about “Fantastic Beasts and How to Rank Them” [1]. As a member of the original International Society of Cryptozoology, as well as life long fan of fictional worlds of all kinds, I enjoyed her summary of recent psychological research on how people think about “impossible” things.

As the title implies, some of this research examines how people reason about imaginary entities and situations. Is a Yeti more or less “impossible” than a vampire? Is levitation more or less impossible than becoming invisible? And so on.

The interesting thing for psychologists is that even though people may agree that something is imaginary and pretty much impossible in the real world, we can not only imagine it, but imagine the world that it exists in.

Of course, imagining the not (yet) real is the heart of creativity of all kinds, so no one should be surprised that people do it. And imagining how a not-yet-real think would really work is the crux of both invention and story telling.

The recent psychological work has worked to tie this imagination with intuitive physics, the unscientific scientific rules that people learn about how the world works. AKA, “commonsense”. For example, objects do not change into other objects. Big objects are generally heavier than little objects. Stuff like this.

Schulz discusses recent experiments that sort through different sets of these rules as they apply to imaginary animals and situations. Essentially, the concept of “impossible” can be broken down into a range of ways that things are impossible. Some things are “impossible” in many, many ways (such as Hollywood vampires or time travel). Others are actually possible, but just not actually factual (such as Yetis or visitors from outer space).

As she notes, at least from what people say in psychology experiments, there is often strong agreement on such decisions. This is very interesting because it offers a glimpse into what and how people learn about the world. These imaginary cases shed light on everyday reasoning about the real and the possible.

As is often the case, psychologists would benefit from walking across the quad to talk with some Anthropologists about this topic.

Schultz hints as some of the cultural variation that can be found in the world. Almost everyone has grown up with tales of ghosts, but the details are different in different traditions. Hollywood has blurred folk cultures with its own super-cultural mish-mash, but Chinese ghosts and vampires are still quite different from English and Transylvanian entities.

These differences are due to the critical role of story telling. Humans like to tell stories,which tie up events, causes, and effects into a coherent narrative. Stories give explanations for random and inexplicable events, and describe the world at a human level.

(Perhaps the key innovation in all of “science” is that it uses a different kind of story, one that isn’t human centered, includes randomness, and is not judged by whether people like the story or not. Stories, yes. But not just any story.)

Many “impossible” animals and situations are known to us through stories, not through experience.  When people are judging Yeti vs.. Dracula they are working from folk tales, not from scientific journals or personal experience. These stories may be based in cognitive illusions (e.g., ideas about disembodied souls) and intuitive physics, but mostly they reflect the motives and anxieties of the society they come from.

Hollywood vampires are scientifically improbable for sure, but some of their features are obviously ideological. Setting aside the evident deep, deep anxieties about the seduction of young women, Hollywood vampires are associated with demonic forces, and are supposed to be allergic to crucifixes and holy water as much as sunlight and silver. These traits is obviously Christian propaganda, painted onto folk tales about revenents. And, by the way, the supposed effects  of a cross on a Vampire is just as plausible or implausible as your beliefs in holy water, crucifixes, and exorcism—which is a whole different psychological question.

Taking verbal reports as indications of folk-science also misses the key point that many such tales have become symbols of specific cultural identity. Endorsing Bigfoot, Biblical literalism, and the everyday influence of demons and angels may be as much about asserting cultural solidarity (or resistance), as a literal claim of truth. This has nothing to do with reason and evidence, and everything to do with personal identity.

I may say that Bigfoot is more likely than Zombies, but maybe that’s reflecting my preference in popular TV shows, and the sub-cultures they reflect.  This belief is social signalling, not pseudo-scientific reasoning about the world.

Finally, I’ll suggest that the psychologists and new anthropologist friends toodle across the quad again, over to the business and law school. Over in that part of campus, the boffins operate in the most dangerous fantasy world of all, one that believes that humans are rational creatures with common sense.

We aren’t. We are fabulists, who believe absurd stories about the world all the time. Any theory that doesn’t take that as an axiom is just plain broken.

  1. Kathryn Schulz, “Fantastic Beasts and How to Rank Them”. The New Yorker.November 6 2017, Conde Nast.


How Zebra Finches Learn A Song

The second coolest things about birds (the coolest thing is that they can fly, fer goodness sake!) is that they sing. Furthermore, song birds learn their songs from the examples of their local group, just like humans learn to speak.

Learning a new song, just like learning to speak a new language, is hard, especially with only examples to work from. In fact, it would seem to be logically impossible—but everybody does it!

This month Dina Lipkind and colleagues in New York and Zurich report a neat study of how Zebra Finches learn a new song [1].

Their basic observation is that learning a song (or utterance) requires learning to make the sounds and learning to make them in the right order and timing. In any but the shortest song, the number of possibilities is far too large for trial an error, or even for progressive approximation of the model. It’s computationally infeasible, AKA, impossible. (The imperfection of the model, and the presence of multiple exemplars only makes things more complicated.)

Lipkind and colleagues explore a possible simple, if non-optimal, strategy: decompose the learning into pieces. Specifically, the student might learn to make all the sounds individually, and then fix up the sequence later.

The experiments use Zebra Finches, a favorite and well studied species. The investigators presented recorded and synthetic songs, and recorded the partially learned songs. Specifically, they analyzed the errors in detail. (See the paper for details.)

Their conclusion is that these birds have neural machinery for learning the pitch of tones (i.e., the pieces of the song), which learns the tone relative to neighboring tones, not the whole song. Essentially, this learns each piece, without trying to keep things in order.

A second phase learns the correct sequence, which involves learning the transitions between the pieces. The researchers conclude that this is a separate neural mechanism.

“Thus, zebra finches break down the computationally difficult task of exploring the entire space of possible motor permutations, into two simpler tasks, yielding a search for solutions that is non-optimal, but manageable.” (p. 8)

This is an interesting finding, because it parallels aspects of human learning. Furthermore, it is analogous to the most successful strategy for computing document similarity. This isn’t so much bioinspired design as bio-confirmed convergent design. Cool!

This study was non-invasive and didn’t directly measure the brain activity. So it is not known how these processes are realized in the finch brains. The study does constrain the types of neural structures that might perform as observed, though. Future research may be able to document the internal processes.

The study leaves a number of questions open. For one thing, how are the individual segments (“gestures”) learned or, for that matter, identified as units. E.g., for human babies, learning to utter phonemes is relatively simple, but only if you know how to find the beginning and end of a phoneme—which isn’t obvious.

I would note from experience that this problem is endemic in learning algorithms. I have struggled with machine learning attempts to recognize gestures from movements. It’s very hard, if not impossible, to induce where gestures begin and end, at least without supervision.

Further studies of songbirds may offer bioinspired tricks for learning gestures from examples. (E.g., hints about what is a gesture to be learned.)

The researchers speculate that this decomposition might be an evolutionary compromise between that offers efficient learning with relatively little neural resources. The intuition is that the two simpler learning processes require significant neural tissue and energy, but much less than more complicated optimization strategies. Proving this case will require a lot more information about how zebra finch brains actually work, and also what conceivable alternative architectures might be.

  1. Dina Lipkind, Anja T. Zai, Alexander Hanuschkin, Gary F. Marcus, Ofer Tchernichovski, and Richard H. R. Hahnloser, Songbirds work around computational complexity by learning song vocabulary independently of sequence. Nature Communications, 8 (1):1247, 2017/11/01 2017.


A Big Hole Inside the Great Pyramid

Pretty much everyone has heard that Scan Pyramids Mission has detected a large void in Kufru’s Great Pyramid [2].

Whoa! Using cosmic radiation scanning a pyramid! That’s heavy, man!

The 4,500 year old pyramid is one of the oldest and largest structures ever constructed by humans, and there is much we don’t know about this artificial mountain.

For one thing, it isn’t easy to know what lies inside a million tons of rock.

The SPM used muon radiography to image the interior of the pyramid. This is a pretty cool technique, using muons that are generated by cosmic rays hitting the Earth’s atmosphere. These tiny particles shoot through the air, water, and stone at near the speed of light, but they are absorbed and scattered by matter. So, dense stone will block more muons that air, revealing a silhouette of the internal structure.

Basically, a detector is set in place to count muons coming down from the sky, through the pyramid and hitting the device. There are zillions of muons passing all the time (something like 10,000 per square meter every minute), but they rarely interact so it takes a while to accumulate a clear picture. The measurements took a couple of years, sitting there catching muons. (There is some clever analysis required to interpret the muon counts – see the paper [2].)

Their initial observations detected the known chambers, and also suggested an unknown large low-density cavity above and parallel to the Grand Gallery. The team confirmed this finding with two other instruments.

The conclusion is that there is something above the Grand Gallery.

The researchers are careful to call it a “cavity”, not a chamber. There are several cavities known in other parts of the pyramid, likely left by the builders to reduce the pressure on the internal chambers. It is quite possible that this newly found cavity has a similar origin. (It is located above the very large Grand Gallery, which must surely be one of the most vulnerable structures inside the pyramid.)

On the other hand, it certainly seems large enough to be a new chamber, and if it is, there could be unprecedented artifacts hidden there—for more than 4,000 years.

The investigators are working on ideas for how to explore the chamber. An initial concept is to drill a small (3 cm) hole, and slip in tiny robots to look around.  Maybe tiny UAVs, assuming the chamber is not filled with sand.


  1. Jonathan Amos, ‘Big void’ identified in Khufu’s Great Pyramid at Giza, in BBC News -Science & Environment. 2017.
  2. Kunihiro Morishima, Mitsuaki Kuno, Akira Nishio, Nobuko Kitagawa, Yuta Manabe, Masaki Moto, Fumihiko Takasaki, Hirofumi Fujii, Kotaro Satoh, Hideyo Kodama, Kohei Hayashi, Shigeru Odaka, Sébastien Procureur, David Attié, Simon Bouteille, Denis Calvet, Christopher Filosa, Patrick Magnier, Irakli Mandjavidze, Marc Riallot, Benoit Marini, Pierre Gable, Yoshikatsu Date, Makiko Sugiura, Yasser Elshayeb, Tamer Elnady, Mustapha Ezzy, Emmanuel Guerriero, Vincent Steiger, Nicolas Serikoff, Jean-Baptiste Mouret, Bernard Charlès, Hany Helal, and Mehdi Tayoubi, Discovery of a big void in Khufu’s Pyramid by observation of cosmic-ray muons. 11/02/online 2017.

PS.  Wouldn’t  “Sitting there catching muons” be a good name for a band?