Jump to content

Welcome to FutureTimeline.forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!
Photo

There is a photonics revolution underway -- and will continue for the next several years.


  • Please log in to reply
13 replies to this topic

#1
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
It seems like every couple weeks I read about some new advance in photonics. Just yesterday, I read about this:

https://techcrunch.c...-head-of-a-pin/

LIDAR you can (almost) balance on the head of pin. This is the kind of thing that might advance the field of brain-computer interfaces, using time-of-flight tracking of light -- a method that Facebook seems eager to build on:

https://tech.fb.com/...-saying-a-word/

Then there's holodeck-like light-field displays:

https://www.lightfieldlab.com/

Then there are all the advances in smartphone screens. For example, microLED is making waves:

https://www.macrumor...future-iphones/

There are now foldable screens, screens with logic layers and embedded cameras, 3D sensors, and so on.

Mary Lou Jepsen's OpenWater is, apparently, making use of some of these new innovations to build their medical scanner / future-BCI, just as Facebook is hoping to use LiDAR-like technology to do the same.

There are also flat, lenseless cameras. In fact, there are several different types. Two types are discussed in this thread:

https://www.futureti...tionize-optics/

(One type is based on patterned nano-structures to focus light like a lens; the other is based on compressed sensing.)

Oh, and let's not forget all the advances in solar power collection. I saw one just recently about some scientists increasing the theoretical efficiency by leaps and bounds, but haven't looked closely at it.

What might we see beyond these trends that are already stated? I'm not sure, but I bet there are many surprises ahead, just as all these above were found to be surprising, and the domain of science fiction not so long ago. I will have to put my thinking-cap on and try to look ahead at what might be possible, and then write something here later; it will largely be guesswork on my part, though, since I am not a photonics expert (but know just enough to make some educated guesses).
  • Raklian, eacao, Casey and 2 others like this

#2
tomasth

tomasth

    Member

  • Members
  • PipPipPipPipPip
  • 243 posts
Could a satellite swarm based photonics , read/write to one's brain from LEO ?

#3
funkervogt

funkervogt

    Member

  • Members
  • PipPipPipPipPipPip
  • 838 posts

Well where is it? 



#4
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
At the moment, mostly in your smartphone.  But soon, also, self-driving cars, robots, drones, medical equipment, power plants (solar panels), BCIs, and maybe home appliances.  It's been building for years, and is reaching new heights.

One thing I will add is that the photonics community is one of the most secretive / closed of any I've seen; they thwart outsider attempts to figure out what they are up to. Machine Learning people, in contrast, love to share their work, post thousands of YouTube videos of talks, have many podcasts, and so on. Photonics papers published by OSA and SPIE are pay-walled, and very few of their conference talks get posted to YouTube. The ones that are actually public are very technical and opaque to someone not in the field; in contrast, Machine Learning folks explain their work in plain English in their writeups. I find it very hard to track down much information about the goings-on at some of the big optics / photonics conferences -- but very easy in the case of Machine Learning.

Look, for example, at Facebook's BCI work, and also Oculus: they kept what they were doing under wraps for 2 years -- hardly a trickle of information leaked!

I guess part of the reason for all the secrecy is the amount of money involved. There surely is a lot being invested in smartphone screens. Then there was the $1.5 to $2 billion spent on Magic Leap, alone; the billions spent on Oculus; the massive solar power industry; and so on.
  • Raklian likes this

#5
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
Ok, here is a radial, slightly crazed idea -- that may not end of being developed, but shows just what kinds of things might be on the (far?) horizon you might not think of:

It is known that certain materials can harden or loosen when exposed to certain frequencies of light. This is used in some types of dental fillings to good effect. But perhaps there is also a material, in the form of a liquid gel -- maybe a mixture of several different molecules -- that can rapidly and strongly expand when subjected to one frequency of light, and can contract (or at least return to baseline) when exposed to another. The amount of light is also important here -- you need not just the right frequency, but also it has to be of sufficiently high intensity.

I'm assuming here that the energy needed to make the gel expand (and perhaps harden) or contract (and perhaps loosen) does not come from the light, but is present in the gel itself; the light only would act as a catalyst or activator of the state change. The gel's energy source might be something like ATP, and would need to be continuously replenished, just as blood continuously replenishes the body's access to energy.

Now, if you have such a gel, you could use it to build robot muscles that could be exquisitely tuned -- to a much greater degree than actual biological muscles.

The way this could work is that you could project holograms ("wavefront shaping") at the expansion and contraction frequencies of light into the gel. Each little voxel of gel would respond accordingly -- some would expand, others would contract. You'd have thousands of degrees of freedom to play with, in terms of how to shape the muscle.

This robot would have no motors or actuators -- just a gel of uniform composition, a membrane around the outside to keep it from leaking out, an LCD array and laser for the holograms, and maybe a way to recycle the gel to keep its energy supply up. Maybe the hologram projector could be in the form of a skeleton, projecting the light from inside the gel itself. This would sacrifice surface area, but you'd still have a lot of degrees of freedom to play with. Note that, since the gel would have a uniform composition, and since you would know the expansion-and-contraction state of each voxel, you would know the shape of the wavefront needed to change its shape; you wouldn't need to use ultrasound to figure out where to focus the light.

I daresay that this would drastically reduce the cost of building high-end robots, and that are orders of magnitude more graceful than existing ones.

One could imagine a walking robot with extremely fluid motions -- and the ability to shape-shift, just like the T-1000 robot from the Terminator movies.

Projecting the holograms would not be difficult to pull off, using modern smartphone screen technology, I would guess. I mean, this is basically what Mary Lou Jepsen is attempting.

The hard part would be at the interface between photonics and materials science, in coming up with that gel mixture. And, of course, the controller algorithms to make the robot move gracefully would require some AI advances.

....

Note that there are light-based 3D printers that do something similar to what I'm describing, but at a slower rate:

https://newatlas.com...printing/58301/

And there is active research into so-called "smart gels" with the kinds of properties I described:

https://pubs.rsc.org...5k#!divAbstract
 

Advances in the field of supramolecular chemistry have made it possible, in many situations, to reliably engineer soft materials to address a specific technological problem. Particularly exciting are “smart” gels that undergo reversible physical changes on exposure to remote, non-invasive environmental stimuli. This review explores the development of gels which are transformed by heat, light and ultrasound, as well as other mechanical inputs, applied voltages and magnetic fields. Focusing on small-molecule gelators, but with reference to organic polymers and metal–organic systems, we examine how the structures of gelator assemblies influence the physical and chemical mechanisms leading to thermo-, photo- and mechano-switchable behaviour. In addition, we evaluate how the unique and versatile properties of smart materials may be exploited in a wide range of applications, including catalysis, crystal growth, ion sensing, drug delivery, data storage and biomaterial replacement.


And now I'll try to think of something a little more near-term...
  • funkervogt likes this

#6
Hyndal_Halcyon

Hyndal_Halcyon

    Member

  • Members
  • PipPipPip
  • 92 posts
If you've ever wondered about lasers and how easy it is to use light itself to manipulate the entire quantum realm, and how easy it is to manipulate light itself using the quantum effects of everyday objects, you'll be doomed for life in thinking biological lasers can solve everything from telekinesis to full-spectrum imaging to thermodynamic manipulation. You can freeze and burn objects simply by adjusting intensity and frequency, or hold objects from a distant with the right interference pattern, or even teleport information across space and time with seemingly no drawbacks.

What I'm saying is, lasers, mirrors, lenses, and quantum dots are not sufficiently talked about because not a lot of people are imaginative enough to see a world of imaginary colors beyond colors we know now. They see street signals but not Wi-Fi signals. The inherent design of our eyes limits what we can do, because this is the limit of nature. Metamaterials will be so important in the far future that our entire biology will be based on them that we will practically live on an entirely customizable set of physical and universal laws.

But then again that's just wishful thinking. To think we can fuck up reality just by knowing how light works is really above anybody's paygrade, but it just takes a little time to replace nature entirely.
  • starspawn0 likes this

As you can see, I'm a huge nerd who'd rather write about how we can become a Type V civilization instead of study for my final exams (gotta fix that).

But to put an end to this topic, might I say that the one and only greatest future achievement of humankind is when it finally becomes posthumankind.


#7
funkervogt

funkervogt

    Member

  • Members
  • PipPipPipPipPipPip
  • 838 posts

 

 

One could imagine a walking robot with extremely fluid motions -- and the ability to shape-shift, just like the T-1000 robot from the Terminator movies.

But unlike the T-1000, wouldn't your robot need to have all of its gel contained in sacs of some sort so it didn't leak out or get contaminated with dust and dirt from the environment? For example, let's assume your robot is humanoid and has a simple metal skeleton that is similar to a human skeleton. To move its arm, it would need to have sacs of gel attached to its metal arm bones, in approximately the same places that human muscles like biceps are. The fact that it would have a metal skeleton and would need to keep its gel contained in sacs would greatly limit its shapeshifting capabilities. 


  • starspawn0 likes this

#8
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
Yes, so not exactly like T1000; but with a sufficiently flexible outer membrane, you'd have a lot of morphological freedom. The membrane could also be given the ability to change color, with millions of "pixels" on the surface; so, not only could it mimic a lot of different object forms, but also appearances. One might also be able to mimic complex textures on the surface -- e.g. sandstone-like, papery, silky, and others.

Incidentally, I discovered that people have already experimented with light-controlled soft robots:

YouTube video

It's still at a primitive stage; but I see no obvious reason why something like what I described couldn't be built some day.

Also, it occurs to me that there may even be gels that can absorb the energy needed to flex from yet other frequencies of light. So, there would be three frequencies: one for expand, one for contract, and one for replenish.

#9
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
Another photonics advance we might see is true invisibility cloaks (or "panels", not "cloaks"):

Current invisibility cloaks are limited by the size of the object and the wavelength of light involved. The problem is that people are looking for solutions where you bend the light around an object; and, unfortunately, such a simple method will have its limits.

There are alternatives -- though, they will require using a lot more heavy-duty technology. Consider, for example, a high-resolution light-field display that projects the scene behind an object, including correct parallax and other features you would expect from an invisibility cloak. So, perhaps an object is placed inside a box; and the box and its contents will "disappear".

What's the problem, and why hasn't this been tried? Well, first you need really good light-field displays, and these are only just now becoming available; you also need a very high resolution, compact light-field camera. And then you have a latency to contend with -- the time needed to compute the correct screen output, given a scan of the scene behind the box. Perhaps using Machine Learning, you can make do with a cruder camera, while also predict a fraction of a second into the future what the screen output should be -- to remove the latency. I think that's possible, though would work better in some environments (e.g. in a forest) than in others (e.g. an complex urban scene with lots of moving objects).

I could see, for example, tanks or jeeps with light-field panels placed around them, with very small cameras, and a computing element with sophisticated machine learning, making the object (jeep or tank) invisible to the naked eye. You might notice a small shimmer, where the machine learning model doesn't predict things quite right. A model for what it would be like is perhaps something like the alien from the Predator films, only the invisibility effect would be even better. You could be standing 10 feet from one of these cloaked tanks, and not even notice.

Perhaps cloaking the tracks or certain other parts would require a redesign of the body; but I'm guessing it can be accommodated.

#10
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
One technology you might have heard about recently is using crystals or glass to do Deep Learning:

https://towardsdatas...es-da151b41368c

As long as you have a way to create semi-transparent, glass-like structures with a non-linearity, you can basically build a feed-forward computer; and perhaps even an easily-trainable "neural network", if you can easily change the 3D structure. So, for example, a little conical glass structure could be "coded" to do optical character recognition. You move it over a piece of paper with letters on it, and light at the other end would focus on areas designated for a classification of the letter.

This is not just about Deep Learning, but computation, in general. You could perhaps add an "amplifier element", to magnify the intensity of the light, and also threshold-out light below a certain level. And then you could do computations of indefinite length, by feeding-back the "output" into the "input".

I would not call this an "optical computer", of the sort people have been talking about for ages. This is a much looser kind of "computer", without precisely-engineered chips. Nonetheless, it would have many uses, particularly in medicine.

For example, you could build a little glass "oracle" that can detect eye diseases -- just hold it up to someone's eye, and look at the light coming out the other end; e.g. if the light is bright on-the-left-but-not-the-right of the oracle, then they are healthy, and if bright on-the-right-but-not-the-left, then they are not.

The oracle would look like just another piece of glass of non-uniform composition; but could be used to diagnose a whole range of health issues.

#11
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts

And there's also talking plasma balls:

 

https://m.techxplore...ole-lasers.html



#12
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
Another area where we might see a lot of progress is so-called "photochemistry". This involves taking a mixture, and then shining light at just the right frequency into it. Certain molecules will respond in certain specific ways to certain specific frequencies of light -- basically, the photon may be absorbed, which will cause an electron in a molecule to undergo a state-change; perhaps another photon of a different frequency is kicked out. This can have the effect of breaking-up molecules into smaller ones, than can then combine with other molecules to produce new compounds. Perhaps through a long sequence of hitting a basic mixture of simple molecules with just the right light, at just the right frequency, you can produce specific, complicated compounds -- e.g. medicines.

You might need a way to separate-out the desired molecules from the other ones. And, here again, light can help with that. It can also be used to separate out atoms or molecules by weight, like a centrifuge. This is used, for example, to purify Uranium -- there's a laser-based method that is massively more efficient than the mechanical versions.

Here's an article on the future of photochemistry, and in particular how to make it work at production-scale:

https://pubs.acs.org...cs.oprd.8b00213

#13
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
Returning to the bit about "optical computers", a little observation: you know like in Indiana Jones-type movies where there is a temple from a lost civilization that seems to act "intelligently"? Well... by encoding computing into the structure of glass (+impurities), you could build a temple that can do some crazy things, yet look very "organic" -- only natural power sources required; materials are stone, sand, water, rope, and glass. For example, maybe you have to hold up an amulet in front of the eyes of a god-statue, and then if it recognizes the amulet, it opens a chamber. You could encode a kind of object recognition for the amulet into a glass cylinder behind the eyes of the statue -- and then if the output is "yes", it would trigger a switch to open the door. That kind of thing.
  • funkervogt likes this

#14
starspawn0

starspawn0

    Member

  • Members
  • PipPipPipPipPipPipPip
  • 1,319 posts
I found a SPIE opinion paper published in January of this year (2019) titled

"Perspectives on the future of photonics: the best is yet to come"

https://www.spiedigi...0501.full?SSO=1

A lot of things are listed, but not many things that are unexpected or exciting (the things I listed above are a lot more exciting). Still, you get the impression that photonics is a powerful beast that is growing in strength and influence.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users