Sabine
patreon
Science News Feb 15
🕑 Added 2023-02-15 16:00:08 +0000 UTC[This is a transcript with references.]
Welcome everyone to this week’s science news. Today we’ll talk about a mysterious laser beam over Hawaii, brain scans in court, a link between quantum chips, self-flying airplanes, better sensors for self-driving cars, how to weave a message into your sweater, the durability of solar panels, and of course the telephone will ring.
Many thanks to all our supporters on Patreon who make these videos possible, especially those of you in tier four and higher. And you can support us, go check out our Patreon page, or support us right here on YouTube, by clicking on the “join” button below.
Mysterious green beams split the night sky over Hawaii last week. The eerie appearance was caught by the Subaru Telescope on the summit of Mauna Kea. The telescope is operated by the National Astronomical Observatory of Japan, and is one of the world’s largest optical and infrared telescopes. Luckily the laser scan wasn’t the beginning of an alien invasion. It turned out to come from an Earthly source, if not a source on Earth.
The first suspect for the signal was NASA’s satellite ICESat-2, which measures the thickness of ice sheets and the elevation of forests. It has a laser, which sends out signals very similar to the ones that were seen over Hawaii, about 10 tenthousand pulses a second, which means that one hits the ground every 71 centimetres along its track.
NASA, however, crunched the numbers and found their satellite wasn’t in the right place at the right time, and that it was probably instead the Chinese satellite Daqi-1. The Daqi-1satellite was launched in April last year. It spends its time in orbit checking out the atmosphere and monitoring Earth’s environment. One of its instruments is a laser to take measurements of atmospheric aerosols and carbon dioxide concentrations, and that’s likely what they saw over Hawaii.
China plans a network of such satellites, probing the atmosphere to measure pollutants, so it’s possible that laser scans will become a more common sight in the future. At least we’ll all understand then how it feels to be a cereal box at the register.
The earth isn’t the only thing that’s been scanned recently, brains are under investigation too, and not for entirely academic reasons. Scientists from the University of Virginia have proposed that brain scans should be used in court. No, not for detecting lies, but to aid in trademark disputes.
Court decisions often rely on whether a “reasonable person” would come to a particular conclusion about something, be it the similarity of two products in a copyright case, or whether there’s actual damage to someone’s reputation in a defamation hearing. British law uses the charming Victorian phrase “the man on the Clapham omnibus” to refer to this to this hypothetic reasonable person. If you find the reasonable person, please tell them to get in touch with me.
In reality, this ‘reasonableness’ test is carried out by judges who may, shall we say, not be particularly representative of the average person. If you think technology shouldn’t stick its curious nose here, let’s imagine you see a toothpaste brand, Colddate with this packaging.
Does this remind you of something? Not that failed date at the ice rink, I mean something, a I dunno, like another toothpaste maybe. Well, a judge said it’s not similar enough to represent a trademark infringement. Guess he doesn’t often take an omnibus in Clapham.
A new paper from the University of Virginia now argues that justice would be better served by using brain scans to measure people’s reactions to different products. Their method makes use of what’s known as “repetition suppression”. If you see a new image, that attracts your attention and creates an identifiable signal in your brain. If you see the same image again, that attracts far less attention. This is why you get bored if the only thing that happens in this video is me talking, and this is why you can use brain reactions to identify how similar two images are. The smaller the brain signal of the second image, the less “new” the image is compared to the first.
The Virginia team developed an analysis of brain signals that is up to the task. They then did some experiments on volunteers and, by checking the scans against self-reports, they found that the brain scanner results are less influenced by presentations of the evidence.
But it'll probably take some time until judges are willing to admit evidence from brain scans, which returns us to the problem that reasonable people stay well away from court.
At the moment, researchers increase the number of qubits in quantum computers by putting more of them on one chip, but this method will ultimately become impractical if not unfeasible. It makes more sense to mass-produce chips and connect them. But the connections between the qubit chips must be able to maintain quantum entanglement, which makes the process challenging.
The new work is a collaboration between researchers at the University of Sussex and the startup Universal Quantum. The qubits they used were traps for Ytterbium ions. In such devices, the ions are trapped by lasers and more lasers are then used to entangle the ions. In their paper, the authors report that they successfully transported the information between qubits with an error rate below one in a million, at a rate of more than 2000 quantum bits per second.
Both of these numbers are orders of magnitude better than previous solutions for ion traps. They achieved this by connecting two qubits not by photons, as had previously been done, but by electric fields. Here you see how it works, it’s a single ion jumping from one chip to the other taking the information with it.
These ion trap qubits which were used for this experiment are different from the superconducting qubits that are used by IBM and Google. Ion traps operate at somewhat higher temperatures, though they still need to be cooled to few degrees above absolute zero. This may not be your idea of “warm” but it’s a far higher temperature than you need for superconducting qubits which is typically a few milliKelvin.
Ion traps also have a longer coherence time, which means they keep their quantum properties better. While the coherence time of superconducting qubits is typically a few microseconds, ion traps can make it up to several seconds. However, they are also slower to operate, and staying coherent for longer isn’t necessarily an advantage if you get nothing done in that time, a familiar problem also for radio interviews.
This isn’t the first time that two qubit chips have been connected. It’s been done many times before, both for ion traps and superconducting qubits. But transferring information with an error rate of less than one in a million is a remarkable achievement, certainly one that I can only dream of.
Hi Joe,
You shot down another balloon?
You’re not sure what it was that you shot at. Okay. Well. Guess that’s alright. I mean it’s America after all.
People will understand that. Sure. Byee.
The Aeroplane manufacturer AirBus is moving closer to fully automating their aircraft and is now testing several new features.
The new suite of automation and pilot assistance features, called DragonFly, will extend the capabilities of traditional autopilots with automated take-off and landing, taxiing assistance, and the ability to automatically divert the plane if an emergency occurs.
They’re really pushing the dragonfly metaphor in their promotional video. AirBus claims to have been inspired by the insect’s amazing ability to recognise landmarks during flight. Dragonflies are known to use this ability to carve out a territory, defend their breeding sites and search for mates. I think we should keep an eye on where airbus is heading with this.
Be that as it may, dragonflies do indeed have excellent vision, and the new system from airbus makes use of advanced computer vision and other sensors to detect where the aircraft is, what the weather is like, and how the terrain beneath the aircraft looks like. It can aid landing by providing a better view of the runway, can communicate with air traffic control, and can even choose the most suitable nearby airport for an emergency landing. In the recent tests carried out at Toulouse-Blagnac Airport in France, an A three-five-zero dash one thousand test aircraft, that’s a big passenger jet to you and I, was also able to taxi around, guide itself to the correct runway, regulate its speed, and provide the crew with alerts about obstacles.
Dragonfly is one of several projects that Airbus is working on with the goal to one day fully automate flying. So you can finally become a pilot and fulfil your dream of being unemployed.
Self-flying planes may be some time in the future, but self-driving cars are already here. A group of researchers from Japan has now developed a new sensor that could make autonomous driving safer.
When future-you gets into a self-driving car and heads out on the highway, your autonomous vehicle may be using a system known as LIDAR to create a 3D map of its surroundings. Here is an example of a LIDAR system from the California company Velodyne Lidar. LIDAR works a lot like RADAR, in that it sends out pulses and detects their returns to work out where solid objects are. But instead of using radio waves, LIDAR is based on laser beams, usually in the ultraviolet or near-infrared. So, it’s kinda like having Superman’s laser eyes, but without the explosions.
I say your autonomous vehicle “may be” using LIDAR because the most famous manufacturer of such vehicles, Tesla, doesn’t use LIDAR, but instead a less expensive combination of camera and RADAR and, according to its CEO Elon Musk “You can be superhuman with only cameras”. Though he seems to be less concerned with seeing other people, and more with being seen by them.
A more complex version of LIDAR is the “Flash LIDAR” which spreads out laser beams by a diffraction grating or a lens. This covers a wider area in a shorter amount of time, but since far less light returns, it becomes more difficult to extract a signal from it. This is a problem especially for objects that reflect poorly which become basically invisible.
But the new device created by the Japanese group has the potential to overcome this limitation. It merges a spot illumination from the classical LIDAR with a Flash LIDAR and does so on a single chip. In this demonstration you see how their system successfully picks up on the dark object on the right, which would otherwise remain invisible.
It's not just interesting if you want to go shopping after several glasses of wine. LIDAR is also used by NASA’s Ingenuity helicopter zipping about on Mars, and to make high-resolution maps or 3D scans of the interior of buildings using drones or tripod-mounted scanners.
Another potential contribution to robotic road safety comes from researchers at North CarolinaState University. In 2020, they proposed that traffic lights get a fourth light in white. The white light would mean ‘follow the car in front’ to human drivers, but signal autonomous vehicles to use their wireless communication capability to organise themselves.
In a new paper they now worked out an algorithm and did a numerical simulation for traffic flow at junctions. A higher fraction of autonomous vehicles generally improves traffic flow, but they found that using the white light in addition has a significant benefit, both by decreasing delay at the junction and by reducing fuel consumption. There were no pedestrians in this simulation, because humans are really somewhat of an annoyance if you think about it, but I suppose the white light phase will also serve to more efficiently run them over.
Hello?
Elon? Is that you?
Sure, let me think. How about Tweet of the Moment? Or. Tweet it. Or. Oh, I know. Twittertweet Symphony.
Happy to help. Love you too.
Researchers from the university of Michigan have found a way to make it easier to recycle fabrics. This would be much needed progress. In the UK, around 350 thousand tonnes of unwanted clothing end up in landfill every year. In the United States it’s more than 10 million tons, and the global estimate is more than 90 million tons. The problem is that there are many different types of fabric, and recycling them can only be done at high quality if the different types are separated first.
The new method works by weaving information directly into the fabric of the, er, fabric, by using crystal fibres that show up under different kinds of light. These fibres are made from plastics, are themselves perfectly recyclable, and bend and refract light so they show up when exposed to the right wavelengths.
This idea is already used in plastics recycling, because different types of plastics react in different ways to light in the near-infrared, which can be used to separate them. For plastic, that reaction is down to the properties of the material itself. For fabric, it requires weaving a new material into the fabric. These crystal fibres are inexpensive and only need to make up a tiny proportion of the garment.
So when NASA scans you on the way to work they’ll not only know you’re late again, but also that your panties are 100 percent cotton.
Scientists have found a way to increase the stability of a new material for solar panels. At the moment, we mainly make solar panels from silicon-based semi-conductors. One of the most promising alternatives is a mineral known as perovskite made of calcium, titanium and oxygen. It’s sensitive to a wider spectrum of light than silicon, giving it a higher efficiency, can more easily be used for flexible materials, and it’s cheaper to manufacture, too. Its biggest downside is that it’s not very durable. Temperature variations degrade the material and so the panels don’t last very long.
Now, an international team of scientists, who just published their work in Science, have discovered a way to increase the stability of perovskite by adding a polymer during the crystallisation process. The polymer works by wrapping itself around the perovskite crystals, providing it with a soft layer that cushions against the stresses induced by temperature changes.
The two images B and E here are scanning electron images, B is a usual perovskite, E is the new production method with polymer coating. You can see that in the non-coated versions there are these dark areas which are gaps between grains of the material. It’s those gaps that cause a problem when the material warms up and cools down repeatedly because the material doesn’t react evenly. Adding the polymer largely closes those gaps which improves the stability of the material.
In their test, the solar cells produced this way showed an energy conversion efficiency of 24 percent, which didn’t drop as the temperature fluctuated between minus 60 and plus 80 degrees Celsius. This is on the upper end of what currently used silicon solar cells achieve, though last year a Chinese manufacturer claimed they’d made it to almost 27 percent. However, these are all values obtained under laboratory conditions so they’re probably about as realistic as photos in cooking books.
Comments
Rad Antonov
Seems like this paper is begging to be covered in next week’s news: https://iopscience.iop.org/article/10.3847/2041-8213/acb704
Mr. Breeze
As for the courts, witness testimony is worse than useless and the detectives are afflicted with confirmation bias.
Mr. Breeze
Aw, man, I thought you were going to knit QR codes into my clothing. My mother-in-law is unlikely to point a camera at me. For the knitters, there are instructions on Ravelry. You my have to be logged in to see it. https://www.ravelry.com/patterns/library/knit-a-qr-code---dos-and-donts
Using scanners to solve IP and copyright infringement cases will probably save time but I think it's also a gross misuse of equipment, time and expertise, unless the courts purchase their own machines and pay staff to operate them. I can imagine the use of scanners creeping further into the judicial system rather than just having results of scans used as evidence. I don't know if this is too far-fetched but perhaps juries might be eventually replaced with scans, the interpretations being given by experts and debated by solicitors - cut out the middle-men in the system. (It's 4:30 and I'm awake far earlier than I want to be so I'm a bit mentally incoherent but at least this morning I've slept already.)
