A single laser fired through a keyhole can expose everything inside a room | Hacker News Reader

A single laser fired through a keyhole can expose everything inside a room

gizmodo.com

398 pointsanigbrowl170 comments

"Seeing through walls" motion detection can be done with Doppler imaging and commodity Wi-Fi today using custom firmware. Hundreds of research papers have been published on this topic. The upcoming IEEE 802.11bf standard is scheduled to add that capability to every router in 2024, https://news.ycombinator.com/item?id=27123493 & https://www.theregister.com/2021/03/31/wifi_devices_monitori...

> In three years or so, the Wi-Fi specification is scheduled to get an upgrade that will turn wireless devices into sensors capable of gathering data about the people and objects bathed in their signals... When 802.11bf will be finalized and introduced as an IEEE standard in September 2024, Wi-Fi will cease to be a communication-only standard and will legitimately become a full-fledged sensing paradigm... tracking can be done surreptitiously because Wi-Fi signals can penetrate walls, don't require light, and don't offer any visible indicator of their presence.

IEEE 802.11bf paper (March 2021): https://arxiv.org/abs/2103.14918

Theory: https://groups.csail.mit.edu/netmit/sFFT/soda_paper.pdf

this one reminds me of the movie Eagle Eye.

Why? What is the desired benefit from this that it's to be put into a standard?

Here's the good news description of it. Kind of unclear why it's so important as to put into every Wifi device though https://beyondstandards.ieee.org/ieee-802-11bf-aims-to-enabl...

From the link you posted:

> Measurements obtained with WLAN sensing can be used to support new industrial and commercial applications in semiconductor manufacturing, enterprise networking, and test and measurement equipment. The standard will also benefit end user applications, such as home security, entertainment, energy management (HVAC, light, device power savings), home elderly care, and assisted living.

Sounds reasonable. I still worry about what it means for privacy in our homes, and for the future with ever more surveillance and erosion of our rights.

I would guess that setting up a WiFi receiver outside of anyones living space but within signal range is not illegal if you are not intercepting and decoding packets, so anyone could currently get visibility into any home with WiFi. Police, neighbors, crooks, anyone. However there is some precedent for police, they are not allowed to use thermal viewers to look through the walls of peoples homes without a warrant. This would seem to be similar. It's been disconcerting that the police have just adopted technologies because they think they would be useful without anyone considering whether a warrant should be required. There should be a law saying that any technology use by police requires a warrant unless it has been specifically allowed to be warrantless.

There are some good uses of a WiFi tracker though, such as an automated health monitor. I'd like to have a continuous record of where I am in my house. I could use it to tell how much time I spend on a computer, how much sleep I get, what I spend most of my time doing, how much exercise I get, where I left some object, lots of things. I just don't want other people or companies to have all that information and I don't want the police or government to be able to get it without a warrant.

I'm not sure why you think this is good news:

"The technology can measure range, velocity, and angular information; detect motion, presence, or proximity; detect objects, people, and animals; and be used in rooms, houses, cars, and enterprise environments."

It seems incredibly invasive. Can some peeping tom just turn on his WiFi base station to track if the next door neighbour he is obsessed with has a guy over...if she's in her bedroom...is in the shower...you get the idea. Creepy stuff!

I'd love to have this for home automation purposes, but I can't be in favour of this knowing what the standard FAANG companies will probably abuse this standard for, let alone scummy ISPs.

I expect dumb limitations like "too many people around you, you must buy a more expensive plan to play this audio"

It will probably turn off your lights when your neighbor goes to sleep.

As something to install and run for yourself? Sounds cool in theory but awful for your neighbors, and what should be a no go generally. Think of areas where liberty and privacy are conditional, or indeed as a fundamental feature that any ISP could turn on by default in their mandatory router/modem boxes?

If it works reliably, it could make it trivial to implement things like "turn off the lights 10 minutes after everyone has left the room" and "turn on the lights when someone enters". Right now you can still do that with presence sensors, but they're annoying and require hardware for every corner of the room to work reliably.

With its band steering capabilities, I'm convinced WiFi can already make an approximation of your location. It can't deal with things like reflection and interference for positioning of course, but it's still pretty close to working already.

Of course, I mean this as something I'd like to turn on and integrate into Home Assistant, not as something that every ISP router should come with (which is what I fear will probably happen).

Time to buy shares in CAT 5 :P

> I'm not sure why you think this is good news:

I think GP meant it's a `good (news description)`, not a `(good news) description`

I imagine people will sell directional jammers to point at your neighbors Wi-Fi that will fiddle with the intensity to keep the SNR just enough to receive packets.

Yeah, I can not see a legitimate use for this technology and much less a reason to put it in every router. Sounds creepy and invasive

Not only that but it entirely silent. At least with cameras or IR detectors you can kind of see where they are. If this becomes wide spread -- you'll have to ask yourself who's spying on you. Up late on a Saturday night at 12Am? You're uptight neighbours might want to know about that. On some kind of disability insurance? Good luck! Now they'll install this in your home (as a requirement) and know everything about your movements: the time you wake up, when you go out, when you have guests over...

Do other people have better walls than what I'm used to? I already know quite a bit about where my apartment neighbors are in their space and what they're doing.

The decision of what gets added to such standards is largely a combination of politics and what is actively being used. Some companies and research labs wanted it to be added and they lobbied successfully for it.

This technology is already usable with many existing routers, though to a lesser degree. The standard is meant to define expectations of how the technology is implemented in future 'SENS' devices.

I guess it could be useful for AR/VR or replace Bluetooth in some location sensing applications (like in stores tracking movement patterns, or monitoring systems for dementia patients living alone).

I dislike how this additional privacy infraction is going to be the norm in a few years.

They will sell you thief protection, alarm if chidren or pets move to certain places, rbnb tooling and so on.

And now that we all have micrphones and gps in our pockets, we will also accept 3d live mapping of our lives.

Because "we have nothing to hide"

"Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should."

Enterprises already use Wifi to track customers in stores (albeit in a very simple way compared to the new capabilities). This will enable stores to track customers to a much higher degree.

With this, the must have product in 2025 will be shields along walls and windows to keep the signals within :-)

To keep your neighbors signals out. While you stop using wifi :(

Needing to drill a hole in a wall separating two rooms in order to feed a cable through it recently made me think about using an array of ESP32 in order to attempt to create a mapping of the inside of the wall, so that I don't hit any pipes or electrical cables by accident.

The walls we have here are of concrete with steel, so the traditional, cheap wiring / object detectors are basically useless.

I wonder if this is possible without dedicating months of tinkering and possibly ending with the need to learn about AI to interpret the data.

How would you even train the AI without building lots of walls of different but precisely known structures and densities?

I would imagine a human can do it post-scan with enough time. If you teach an AI to do it in real-time, there’s no need to build a bunch of walls.

luckily, existing walls of known construction are widely available

You'd be surprised at how many liberties contractors of various pay grades can take behind drywall

Seems simpler to just stick a pencil in the first hole and see if you are clear to the other side.

"Wi-Fi is among the most successful wireless technologies ever invented. As Wi-Fi becomes more and more present in public and private spaces, it becomes natural to leverage its ubiquitousness to implement groundbreaking wireless sensing applications such as human presence detection, activity recognition, and object tracking, just to name a few."

"...and private spaces, it becomes natural to leverage its ubiquitousness ..."

Are these researchers soley focused on their next grant only, do they live in another world? Or do they sometimes pause and think about the ethics of their research?

They think about it, they just consider it somebody else’s problem.

The genie is somewhat out of the bottle already, because existing access points can track the position of wifi devices and deduce a lot of this information. This takes it one step further to also track people without wifi devices.

This will be deployed in 3 years or so. I think we need a legal and technical approach that will allow "stand your ground" type of rules. Against my neighbor WiFi waves and FFT extrapolation of what division of the house I am in at the moment and with who.

> The genie is somewhat out of the bottle already

Very much so. As early as 2012, maybe 2013, retailers were piloting technology that would offer guest wifi networks, then track you using the in-store surveillance cameras. Don't even need facial recognition, just tie the person's identity to the phone they logged in with. The camera adds better location data than you could get with the wifi alone.

This was about the time that Amazon was offering a Black Friday discount if you used their app to scan the barcode of a product in a retail store, then bought online — and back then, only checkout was https, so if you were in the data path you could detect that sort of a lookup.

I recently traveled and used JFK's free wifi. I went to a duty free shop to buy some chocolates and now all online ads ar showing me the same woman perfume ad.

I've thought about that.

But what's the alternative? Someone else would do it anyway and use it against you.

Same argument goes for general AI, weapons and everything else.

Even if they do this type of research and don't release it to the public, you can bet nations states already are doing the same research in private, and using it against whoever they want.

Better to have this in the public eye, than to be too scared to carry out the research in fear of others abusing it. because whether the research is done publicly or not, the "bad guys" are probably doing it in private.

Your threat model for ubiquitous cheap sensing tech is nation states? I'm not overly worried about being spied on by nation states. In general, my feeling is that if a nation-state decides I'm important enough to deploy hardware against, I have no real defense.

Now, big tech companies putting it everywhere, and either convincing the stores I shop at or my neighbors to pay for the hardware and slurping up the data - that is a real danger case. And those companies build off academic research like this. Even trillion-dollar companies don't do as much of this research inhouse as they do by sponsoring grants like this.

We can already see where in the store you are.

But your clerks aren’t going to record every step i take for a decade inside your store and run algorithms to manipulate my brain into optimizing store outcomes.

They already do this using security cameras. How do you think the manufacturers know which shelves and the placements of them to bid on?

Also by tracking your phone's bluetooth and wireless radios! It's a bit fuzzy but useful enough to be generally actionable.

This type of placement scoring is almost never done via security cameras. The CV and data storage costs required for that to work at scale would be enormous.

It’s much easier and cheaper to just record what shelve/height/area products are placed in and reconcile that data with sales. The data will be more noisy, but you can capture so much more valuable information for the same investment. You can also usually apply the models in retrospect, since most larger chains already record where products are supposed to be.

Cameras are really only used at corporate test stores in small numbers. Camera data is usually looked at manually and only focuses on a specific product or area.

Relevant: https://en.wikipedia.org/wiki/Planogram

Absolutely they are. Nordstrom and Home Depot got called out for doing it 8 years ago - https://www.forbes.com/sites/petercohan/2013/05/09/how-nords.... The technology has progressed significantly since then.

I've worked on systems designed to send in-store customers custom, extremely time limited (< 1 hour) coupons based on where they are in the store and where they linger. Linger time being used as an indicator of purchase intent for a specific item. This does require that the customer has the store's app installed and running.

I remember seeing this demo five or six year ago. Not a theoretical or startup, but being actively tested in large retailers.

Uhh.... we already do this:

https://dnaspaces.cisco.com/

The Target app has very convenient in-store routing on a store map. Turn-by-turn directions down each aisle to bring you to the Kleenex.

>Your threat model for ubiquitous cheap sensing tech is nation states? I'm not overly worried about being spied on by nation states. In general, my feeling is that if a nation-state decides I'm important enough to deploy hardware against, I have no real defense.

Nation states worry me precisely because I'm uninteresting. They wouldn't deploy hardware against me, but I'm under no illusion that they wouldn't use every mass-surveillance option available to them, including mass exploitation of commodity hardware.

I agree wholeheartedly with the rest of your post.

What reassure me is that the americans are so aggressive and dangerous that they become huge money pit target, leaving the rest of us quite free from foreign espionage.

care to elaborate on this?

I can't tell if you mean that most nation state espionage resources are spent keeping a watchful eye on the U.S. since they see the U.S. as aggressive and dangerous, or you mean that literally American citizens are aggressive and dangerous (i.e. perhaps due to prevalence of firearms with non-conformist, anti-government, anti-authority views), that the U.S. spends much of its own resources spying on its own citizens?

The problem is not that they can't spy on you as an individual. The problem is that they can spy on everyone all the time. And then later use an arbitrary amount of historical data to do whatever they want. I'd rather not see that any government becomes omniscient about the past.

Nation states was more of an example. "Anyone that isn't in the public eye" is what I should have said.

Stuff used by national governments eventually trickles down to the local level.

I’d hate for my town counsel to spy on me simply because they can with off the shelf hardware, that the military developed and deemed obsolete 10 years ago.

Wouldn't the same argument work for a research on, say, cheap and efficient biological weapons?

>Are these researchers soley focused on their next grant only, do they live in another world? Or do they sometimes pause and think about the ethics of their research?

Should we not do any third party security testing or vulnerability detection either?

Ignorance is not a solution to a potential problem but simply means you won't know about it until it's far too late.

Nice funding scheme you have here. Develop a threat, write it up as if you're researching it for prevention. The more you work on it, the bigger the threat will get, the more money you can ask for to research "prevention" in the future. This is literally how a lot of this sort of research works.

It's a digital arms race. You can't prevent it from happening by condeming those who are participating. These unfolding events are merely organic consequences to overarching dysfunction within society.

Your theory is that nobody but researchers looking for funding to publish about threats would find these threats? That seems very optimistic to me.

Seems like a convoluted way to secure research funding. If you don't care who you hurt it would probably be simpler to just go work for a defense contractor and drop the publication pretense altogether.

Research ethics are a very important consideration, but in this case this research is identifying a real risk posed by people's widespread deployment of Wi-Fi. We need to know the risks it creates so we can make our choices appropriately - otherwise the only people who know are criminals and nation states.

>Are these researchers soley focused on their next grant only

Yes, they are. Anyone who is saying otherwise is either gaslighting you or speaks out of their ass. The way science is funded right now is incredibly broken. This is admitted both by people who worked in research during 60s-70s (when funding worked differently) and by honest people who worked in modern academia.

Alan Kay speaks about this often:

https://www.youtube.com/watch?v=wdHtYW_wcAs

Dark Horse podcast spoke at length about the current system:

https://odysee.com/@DarkHorsePodcastClips:b/corruption-in-sc...

I worked in Computer Science (CS) research for 20 years. Some researchers take privacy concerns seriously, many do not and see it as out of scope or a nuisance. CS is still mostly male and hence most research is done by male grad students. Young males are often fascinated by spy stuff (I know I was) and will orient their research around it. The conferences where they publish papers also reward them for it by publishing their papers, which indicates an acceptance of this kind of research by the community. Similarly I saw lots of grad students who arranged their research around quadcopters. They liked quadcopters and wanted to play with them so they looked for research that could be done with quadcopters. If they could do CS research that involved looking at porn and get it published they'd probably all flock to that. So part of the explanation is human nature, but the fact it is accepted and published shows there is little real critical review of potential misuse of this kind of research. It is also really difficult to invent anything that can not be misused, especially in research involving sensors. A WiFi sensor can monitor your heart rate for health purposes but can also monitor your heart rate to sell you more stuff.

However I've been thinking the same thing for many years, that much of the CS research seems aimed at spying on people. Companies that will be hiring those grad students may also be partly to blame by hiring students who do such research, probably because it fits with their own corporate data collection and analysis practices on their customers. Grad students want jobs after they graduate, and developing spy stuff gets them jobs. The military -wants- spy stuff so they encourage this kind of research.

So the solution seems to be to somehow get more awareness of potential misuse factored into paper reviews, encourage more research on mitigating misuse of new technologies, and getting more women into CS. Unfortunately women seem to be leaving or not choosing CS more and more lately. Some women may like spy stuff too even though it has historically been more of a male thing, probably due to the military. I'm not sure there is anywhere yet that you could publish research into mitigating misuse of new technology (except privacy research journals). Sometimes there are solutions for classes of sensors, like cameras or microphones. For example, f you have an interesting application based on cameras but recording video would be very privacy sensitive you might be able to only output the processed result and throw away the video data, and a commercial version could be a chip that does exactly that, making misuse of the video impossible.

Wide-band (and ultrawide-band) microwaves were already used for such through-wall sensing. It was just a matter of time before somebody noticed that there is already wide-band microwave transceiver in everybody's home.

I'm sure you recognize the motivations of people who would investigate technology that can peer into a closed and locked room.

I recall the first time I read about using a laser to image the contents of a room. It was old-fashioned military technology when I read about it then - the 1980s.

> Or do they sometimes pause and think about the ethics of their research

Obviously they don't

This resonate me, but also it's impractical to expect something achievable with tech to never happens as long as humanity exists. It's not even about who invent it. The only solution is new tech to counter... new tech, it's no way back, I believe.

As someone who worked on a DoD funded project, that's how you get the grants.

A very large proportion of physics, chemistry and engineering is funded by defense interests. Academic scientists span a large range of politics, from peace-loving hippies to pro military hawks. Many scientists just run on a heavy justification treadmill ("if it's not our group, it would be another one" and "this tech is not yet ready for deployment anyway" and "we'd be doing this work anyway, so it doesn't matter if we spin it as defense related").

Frankly, ethics is not really a strong point for physical scientists and is kinda selected against by the hiring system. Being able to take DoD money and sleep at night helps a lot to get tenure.

It's the sort of thing where "if we don't do it, someone else will". These sorts of things are better to research and make known to the public so that they can be legislated etc etc.

I'm confused, doesn't commodity wifi just mean the normally used ones? If so why should we wait till 2024?

It's more invasive than that. I've seen MIT research using several WiFi routers to detect heart rate and respiration[1]. It's not too much of a leap to using it as a lie detector. A little more of a leap, and you get cameras watching window shoppers, plus some Machine Learning, to get a system able to detect which items excite which shoppers when they look at the items (via heart rate and respiration). Could also maybe be used to detect shoplifters.

[1] https://www.medgadget.com/2014/06/mits-wifi-system-detects-p...

They can't get a lie detector to work with the sensors strapped to you...

Does it really matter? The more ambiguous the result, the better for them.

For sure. If people want something enough, somebody's going to try to sell it to them whether or not it works. People really, really want lie detectors.

My experience in having a polygraph test applied to me after the business I worked at was robbed is that the results absolutely have no bearing on how the cops act afterward. They only bring you in for that in order to tell you that you failed, and should confess now in order to secure a lighter sentence.

Had I not had a family friend who was an ex-police officer, I would've been absolutely scared shitless during that experience. Even knowing they were going to do it, I was still absolutely terrified.

It's a fucking travesty that they are allowed to use pseudo-science mumbo-jumbo to terrify people into confessing. I lost all faith in the US police system in 2001 when I had to go through that. I have zero faith in their ability or even willingness to be 'the good guys.'

For who?

How much existing advertising technology works seems to be questionable. Lie detectors aren't admissable in court for good reason, but they are still considered useful to help lead investigators in the right direction, especially against people that aren't trained or trying to defeat them.

Mostly because of the placebo effect ie most people think lie detectors work and it makes you uncomfortable and stressed to have a device strapped to you while someone with authority asks serious questions.

It leads them down random directions because false positives are extremely common and more importanty false negatives can lead an investigation away from the truth.

It picks up one set of signals that can be altered based on the interviewer. And is colored by the person trying to understand the test.

a stress detector that merely alerts security to take a closer look is a lot more achievable.

It's great for healthcare to general public. Think of cheap non contact heart rate monitor.

Slays me that wifi ended up with IEEE instead of the IETF for example. The profit motive really seems to drive a privacy-second approach to some of these standards.

RF engineering is hard. IETF is ... scope averse, change averse, inertia-driven, and ... not exactly a bastion of hardware manufacturing excellence. IEEE is also way better than, say, the ITU, whose resistance to packet-based switching is famous and whose systems in that space are largely semi-mythical to extinct and who are politically inclined to actively stand for centralized and billable systems in lieu of decentralized local communications.

> tracking can be done surreptitiously because Wi-Fi signals can penetrate walls, don't require light, and don't offer any visible indicator of their presence.

Maybe lead paint and aluminum siding will come back into style.

Not lead, but same idea [1] It takes a few coats and proper grounding.

[1] - https://www.google.com/search?gbv=1&q=emf+shielding+paint&oq...

How big do windows / glass doors need to be before that is more or less compromised?

I don't have a specific formula to answer your question, but there will be leakage regardless of size. This can be partially mitigated with metal screens or fully mitigated part time using steel scroll shutters or steel door shutters. Steel scroll shutters have the benefit of adding R/U insulation rating in cold weather and in some cases can improve physical security.

We want to be able to get natural light and air into the house. I can make a safe room if I need to, but this is an always-on threat and we can't respond with whole house unhealthy environments. Grounded metal (copper or aluminum) screens should work quite well if designed for the application I would think. But that means 100% screening, no fully transparent openings. Maybe we can come up with fluid filled windows that provide the required shielding. What a PITA.

You can get window film that's RF shielded. It's weird to be in a SCIF with windows, but a welcome change from the norm.

Example vendor: https://www.signalsdefense.com/rf-window-films/

Coming soon:

TV providers that charge based on how many people are in your living room.

Peeping Tom spy devices that tell you when your neighbour is in the bathroom (upgrade to Pro to increase the resolution and find out what they’re doing)

Anti-alibi logs that prove whether you were or weren’t alone at any given place or time.

Industry certification schemes promising to turn this feature off but totally leaving the door open to anyone with a warrant who wishes to turn it back on.

Sooner than you might think!

https://www.independent.co.uk/life-style/gadgets-and-tech/pa...

Wow, soon we won't even be able to invite friends over anymore without the god damn copyright industry taking issue because not enough rent is being paid.

>Anti-alibi logs that prove whether you were or weren’t alone at any given place or time.

This actually gives me the most concern. If it's easy enough to create these logs of location and travel via technology, then it's easy enough to fake these logs. And, when the technology isn't questioned in the general population, you are just immediately guilty in the court of public opinion, let alone in actual court.

"PLEASE DRINK VERIFICATION CAN TO CONTINUE"

https://i.imgur.com/dgGvgKF.png

You know, you could put a simple circuit across the opening, powered by one of those extremely low power solar cells, handy RFID chip ... they could make sure you are drinking your verification can.

So how do I jam these signals?

Besides not using WiFi? If you want WiFi to work on your phone while lying in bed, then that requires Wifi signals in your bedroom. Also, remember that they are putting "smarts" into every appliance under the sun, including tea kettles. It is going to be a race to sell information about users from every manufacturer. Even if you go to zero smart devices, are your neighbors going to be less enlightened?

I don't know. I just want countermeasures against the dystopian technology we'll be facing in the future. If some device tries to image me or my home without my consent, they shouldn't get any useful information.

Maybe a signals jammer? Maybe some wall material that blocks signals outside our homes? Custom firmware for our WiFi access points to ensure they aren't doing this, or ensure that they do it in our terms and for our own use instead of some surveillance capitalism nightmare?

Faraday cage your house with anti emf paint and stop using wifi.

I'm torn by both how awesome this is and how creepy it's potential is.

First: I think this comment should be a daily post on HN - a staple mantra of sorts, until it dents the stolid bonehead of collective consciousness. Glad it's top comment, even if the lasers got less light...

Secondly to everyone; my admittedly and hopefully incorrect perspective deduced from such examples (there are many) is that there is a quiet, surreptitious attitude in elite tech that if there's adequate intelligence to get away with something profitable, it's unconditionally fair to do so - so long as it's not explicitly illegal in a way that can't be bypassed by any conceivable means. I immediately think of a common tendency for (most?) humans to justify the often grotesque treatment of animals in some industrial environments, which seems based on, rather than any baseline of decency, the intelligence of the creatures vs humans and final satisfaction of the debateably necessary end product; eg the same exact actions by an individual done for recreational purposes would be a felony. But I strongly suspect this attitude readily crosses the species gap and much of humanity is viewed thusly. They're dumb, vulnerable, ignorant, defenseless, etc, so let's squeeze em for all we can!$$$! Who wants to be millionaire?

Because I perceive much of modern tech as exploitative and even parasitic, I wonder what a comparison between pretech and modern tech exploitation might reveal about the LTS version of ethics vs beta. I suspect that historically, exploitation was more kinetic, unambiguous, conspicuous and effort oriented, whereas modern versions are more ethereal and can take place remotely on an LCD screen pretty passively. In the past, a person typically required persuasion or proximity to make divulge unamiable information. Now, it is involuntarily invisibly siphoned and reapplied to them often undetectably, systematically. I suspect a novel form of unchecked pathological detachment in the ever-growing tech industry, when billions can be rapidly affected by simply executing a blob of effective code, and the consequences reaped as bank deposits, preferred coups, mass behavioral changes, etc. In this environment, even good-natured people can do bad natured shit while keeping ostensibly clean hands and feeling dignified - it's detached, and it's a job. Though I fear a serious diversion between what ought to be more parallel vectors of ethics and development/advancement. A rapidly accelerating diversion. Would this newfangled wizbang be more fugly if viewed holistically and are we weaving a labyrinth of inextricable connections that will only make less and less sense if viewed so? Is there really a future in merchandising every aspect of ourselves, selves that are targets themselves, intended to be corralled into market categories of exteriorly defined self?

There are obvious complexities with concepts mentioned, in a nascent world where so many trades and crafts are becoming obsolete and people need to survive or thrive by any means necessary, increasingly digitally so. There's actually great incentive to produce really clever but dubious shit. I can't help but see an epoch of anomie, of unprecedented fundamental changes in self perception and values, a new human resulting. I don't presume to predict the consummation of this, but sure wouldn't mind skipping the median phase. I'll now tuck my cauda and hide, but with a question:

We see that as modern technology develops, the individual is and has been adapting to decreasing influence of presence and control of it, both in personal and external application. If this trend continues, what are we getting ourselves into? Are we leaping into a collectively agreed-upon, consensually interpreted greater humanity that I wasn't informed of? Or are we a moth in deep space, fluttering toward a new light every moment with no shortage of singularities?

We are now in an undeclared technocracy, veering toward a resurgence of global public acceptance of autoracy/fascism, while people watch TikTok.

Considering recent research linked the breakdown of insect populations to light pollution, that's probably not such a bad analogy.

You assume that conscious ill-will is required. The power and problem with human organizations is that they can achieve monstrous things while ignoring the results of their actions (see: biodiversity loss, air pollution, plastic pollution, nuclear weapons, etc.) while no single member or group takes ethical responsibility. Such failings are well studied: see academic literature in areas like mass psychology / anthropology / organizational psychology.

If I implied "conscious ill-will" was required, it was unintentional, though I do assume there's more than an ample amount to marvel at. Certainly some nasty roads are paved with benign intentions. I also didn't intend to attribute all human failings exclusively to tech, though it does seem wantonly irresponsible in many areas. Regarding autocracy and fascism, I see them as elements in a far greater stride toward totalitarianism. I'm also not convinced that consciousness is saint of the ignorant. Perhaps many are truly unconscious, but if it's self-imposed, it's not inevitable. One point I single-fingeredly (typing on a small phone screen) attempted to indicate in my ramblings is that the remoteness, impersonal, quasi autonomous aspects of tech and the digital era make detachment much easier. The more manifold, indirect and distant the effects our our actions, the easier, but not necessarily justified they are to ignore. It's much easier to pilot a drone to explode dozens of civilians than it is to peer into their terrified eyes while slitting their throats one by one, with one's own hands. Yet the result is essentially identical and the most average of persons is capable of understanding this.

It seems strange that it takes the dilemma of the unvaccinated to consider ignored responsibilities. The majority seem under the impression that by remaining unvaccinated, one is directly terrorizing humanity. Maybe so; but why is this such a lonely example of passive behavior as such a threat?

I don't think I'm assuming much, really. I think I'm simply perplexed by the horde of elephants in the room. Thanks for the reply.

Really interesting how 1984 actually underestimated the level and capabilities of invasive tracking tech. Instead of having the government force it on us, we just happily install it as the next new upgrade.

I'm curious if any of the dystopian sci-fi works actually thought that, instead of the government forcing down and subduing a potentially unruly populace, that they'd basically use "creeping brainwashing" to actually get the populace to demand more top-down control (e.g. how so many Chinese internet users are vehemently pro-government).

Reminds me of “I'm Not Supposed to Show You This ” – Edward Snowden (2021) - https://news.ycombinator.com/item?id=28361341

IEEE 802.11bf will happen, because we allow it to happen and after it has been established everywhere for some questionable reasons everybody is looking like a suprised Pikachu when stuff like this is being abused.

While this seems to be a neat thing, I am just wondering how common nowadays are door locks that actually have a hole through them? Where I come from, those locks could be seen in doors of old cabinets and other furniture. Or perhaps at the summer cottage, but even there, it is usually just an old shed or something; any doors that are supposed to keep anything valuable safe, have solid lock housing, i.e. no through keyholes.

Most of the locks I've seen in my area still have a path for light to bleed through, but it's very, very tiny. Most likely not enough for this application. But peepholes are still a thing, not everyone has a cover on the apartment side, and the effects of the fisheye lens in the middle could be compensated for with a bit of extra math...

You probably don't need even need math, another lens might do.

"Keyhole Imaging" is only a catchy name for the technique. It works with any other opening, like a door crack for example, as long as you can shine a laser through it. I would expect that it is even possible to use an opening that is covered by a light diffusing material like paper or tape on the inside. Instead of measuring reflections from the far wall, light diffused by this material would be used instead.

Same thought. The only house I had with an actual keyhole was built in 1948. Every other house I've lived in has had no such thing.

"can expose everything inside a room"

Can estimate the size, position and movement direction of moving objects within a room.

And shape… based on the recovered data this would be sufficient to say figure out the position of hostiles in a room in a hostage situation for example.

How do you know whats a hostile and whats a hostage? I think that was a movie trope too of dressing up hostages as hostiles and them getting shot by their rescuers.

The images they recovered were clear enough to see who’s holding weapons…

So the hostiles in this case would just tape guns with empty clips into the hostages hands.

Real life isn’t a movie… and even if it was hostages don’t move like hostage takers do.

"We got a blob at 2 o'clock that's holding some kind of a Kalashniblob and in the other hand it looks like either a large donut or a beret of the National Liberation Front."

There is a huge market for wall scans (for non-intrusive wire routing) and a huge market for automatically generated floor plans and structural analysis (for real estate and renovation markets) and yet there are very few products that are effective in this space, which leads me to believe that the technology here is very much proof-of-concept and rarely if ever useful for anything other than a toy for research.

If you can do keyhole scanning of room layouts then productize it!

This technique only works if the object inside the area is moving, and if you can shine a laser into the space in question.

*if an object is moving a crude shape of the object can be made out

"seeing inside a closed room" is a bit of a stretch

That's quite a lot more than not seeing anything. Could be useful if your goal is to determine the number and type of occupants in the room before entering it. I could imagine all kinds of mischief bringers being interested in this.

As a bonus, if your keyhole is wide enough, I guess you could image static objects by moving the laser around.

Or you can just jam a borescope through the keyhole and see things that way. I don’t think peeping through a keyhole into a room was a hard thing even before this.

Watch the video. They can also determine the crude shape of a non-moving object.

I don't think it's a stretch. You are seeing inside it, just with extremely low fidelity but the headline made no claim about the fidelity.

Something like this could be really useful for police or armed services who are about to raid a building. I'd wager they'd be the biggest sponsors of taking this study further.

Prior work, https://news.ycombinator.com/item?id=27124691

I'm confused by this. You can buy a 1.8mm borescope online for $1k. I imagine professional spy organizations have much thinner ones or fiber optic options. So if you already have a hole, how is this helping? Dark rooms? Or an instant room layout maybe for breaching?

I think it's more of a exploration of what information is possible to gather from just scattered light with time of flight measurements and a lot of computation.

It has little to do with "keyholes", other than to give readers an illustrative idea about the experimental setup.

Very interesting and also impressive. Intuitively, one might at first think that "seeing inside a closed room" implies "...without being seen yourself", however that is obviously not the case, since occupants of a room will hardly miss a giant beam of laser.

But then again that's not the point of this technique, and I can well imagine it being useful in rescue scenarios and such.

Would the technique not work with a laser outside the visible spectrum?

I expect this would work with infrared lasers as well.

In which case a standard radar/laser detector made for cars would alert you, correct? [0]

Tangentially, I always loved the story of the US Navy using consumer automotive radar detectors in fighter jets. [1]

[0] https://copradar.com/chapts/chapt5/ch5d1.html

[1] https://apnews.com/article/ff8979c9f5b1b9eb0561a97a6a0af387

Laser detectors might be hard to find in living rooms but you could accidently detect it with the selfie camera of your smartphone. At least in the past, manufacturers cheaped out on this one and omitted the IR filter on it. You can try if that's the case by pointing an IR remote control at it from close up and see if it lights up when you press a button.

It would be hard to detect a microsecond-long laser flash even in visible light, never mind infrared.

The technology is neat but I can't help wondering who has a keyhole anymore that you could actually shine a laser through?

I mean, they could probably just as easily use a window.

My sci-if part of my brain tells me that in 10-20 years we’ll see small box-like devices that you attach to a key/peep hole or perhaps with a built in silent needle puncher that creates a small hole on-demand. Perhaps with a moving laser that would imagine still objects too.

Combine that with AR glasses and you would have a useful pre-breach tool for police/military.

They will just silently access the brain interface of one of the people in the room and see with their eyes instead. If the suspects are lucky, the police will even get a warrant for that first.

I’m sure something like this exists somewhere already.

I bought a tiny camera on the end of a 50’ coax cable to fix a plumbing issue on Amazon for $40… that tech was James Bond stuff 30 years ago!

Did it help with the plumbing issue?

As far as I know, the military and SWAT already use boroscope style cameras with microlenses to see inside pre-breach. I'm not sure this is an improvement

Boroscopes require you to have physical access and in some cases drill the hole.

This can probably be developed to be a remote sensing solution with quite a bit of range and with a non visible it won’t be noticeable to adversaries.

This single laser technique also likely has implications for other NLOS imaging situations too for example imaging a force behind cover by illuminating an area within line of sight of both you and the target.

It would be cool if you could just shoot the plug into a building and it would bore in and start collecting data. You could fly a drone over town with one of these guns and shoot a probe into every structure.

I watched the embedded video, and I see how this could plausibly give you the distance that the photon travels between the laser and the object, but I don't see how you learn anything at all about its direction? Wouldn't the timing look identical for all points that are the same distance away from the point that the laser meets the wall?

It requires the object to be moving. Presumably the shape of the burst of photons returned from a pulse changes as the object moves. It's obviously not simple time of flight distance measurement since that exists.

Impressive! Their technique reminds me of Dual Photography[1], except using the target object's motion over time instead of multi-dimensional lights ("projector") and/or sensors ("camera").

I wonder if these techniques could be merged? Perhaps using this type of 1D sensor over time instead of Dual Photography's 2D camera?

[1] https://graphics.stanford.edu/papers/dual_photography/

Now...all we need to do, is find doors with old-fashioned keyholes...

Come to Germany, there are loads here.

When I first saw keyhole I thought they meant peephole as in the fish eye lens. Next paper perhaps!

(edit: this is still amazing of course!)

I thought keyhole satellite. Was disapointed to read they cant see into closed rooms from space ... yet

..disappointed? Really?

Absolutely, I want to hear about the technological breakthrough that made it possible. Control of spacetime, quantum something, just anything cool.

Kneeling down and peeping through the keyhole was not the answer that I expected when I clicked through.

Fair enough. I suppose I was surprised because I wish there were some things that were just impractical to do. Stuff like this gives me the creeps: https://web.wpi.edu/Pubs/E-project/Available/E-project-04250...

Still, I guess the science is interesting. I'd like it more if it was about eg, noninvasive diagnostic procedures rather than spying.

All agreed - the line between interesting science and creepy use of said interesting science is very fine. That said I have long given up being outraged that carrying around a tracking device makes it possible to be tracked.

> but when paired with a properly trained image recognition AI, determining that a human (or human-shaped object) was in the room seems very feasible.

I am not looking forward to some poor soul being convicted by a jury based on this “AI” showing his real photo

So then, what’s the countermeasure to this? It’s probably already being used for nefarious purposes. So how do you counter it?

you need to wear a faraday suit.

Why not just enclose your room in a faraday cage? Maybe you won’t have WiFi but at least other people’s networks can’t deep into your space

Would this be helpful in any way for fall detection? Especially for the elderly. I wouldn't want a camera watching my parents for example.

I think so, if it can track acceleration it may be able to do so. Would have to play around with it to know for sure.

There are fall sensors the elderly can wear, they often have a bonus panic button too.

In answer to your question: This isn't useful for anything yet as it is early academic research that mostly just proves it is possible, it would have to be refined and minimized to become commercially useful.

That being said I cannot imagine this being more accurate or cheaper than optical sensors + computer vision, privacy issues can be mitigated by building a closed-look with no outside connection. The point of this laser technique is for situations where you cannot/don't want to have direct vision of the location (e.g. inside the body, inside a room, etc).

I guess more generally, a point beam can expose everything that's directly accessible by raycasting through the numerical aperture of a hole without reflections, so long as you can alter the angle of the emitter.

There are usually some space to move the laser around, even if you are restricting yourself to a keyhole. Would that not be sufficient to detect the location and shapes of objects related to the movement of the laser?

I remember seeing something similar but with audio instead of light. The reverberations of a sound bouncing off objects allowed for the mapping of an enclosed space.

Quote: "The research could one day provide a way for police or the military to assess the risks of entering a room before actually breaking down the door and storming their way inside, using nothing but a small crack in the wall or a gap around a window or doorway."

This is ladders and walls situation. And for this technique to be rendered useless the antidote already exists. Criminals all they have to do is start using vantablack - https://en.wikipedia.org/wiki/Vantablack

Or just put tape over the keyhole when you're in the house.

That’s why Cylons only have one laser beam.

No doubt this will be used in military and surveillance long before society sees any benefit.

Misleading title. It can't see _everything_. It only works for dynamic objects.

TLDR; what the images look like: https://youtu.be/Veo27qhrI20?t=193

Pretty cool proof of concept!

Reply on news.ycombinator.com