Don’t let the “quora.com” make you skip this one. It’s a detailed and informative discussion of the U1 and UWB as understood today. I was pleasantly surprised.
It started off sort of interesting (aside from the "people who didn't believe me are redundant bit"), but goes completely kooky by the second half, drawing absolutely absurd conclusions. Or maybe I'm just redundant.
I didn't let that dissuade me, but I felt a little unsure about what I'd just read when I scrolled to the bottom and saw the person who wrote it describes themselves as an "alchemist and metaphysician".
Back in the 2011-2012 era of Square, Brian would often opine at length about our products, our strategy, etc.
He'd write as if he had some sort of inside information and was a serious authority on the subject.
Anytime he'd make a new post about us, it would make the rounds on internal lists. Usually to the tune of "Oh wow, another Roemmele..." and we'd all read on in amazement at how much he'd just plain made up.
As insiders, we knew his posts were pure speculation, and were rarely even close to the mark about strategy or the technical details of our products.
Since then I cringe a little anytime I see one of this posts show up up. He completely lacked credibility when talking about Square. So I'm not sure why it would be different with Apple.
There's a human hack where once you predict/guess/use insider info to get one thing right before hand, you will get some percentage of people who blindly follow you and can explain away any and every incorrect future prediction. I've seen it happen from QAnon all the way to "insiders" "leaking" information on what would happen each week on Survivor.
There's a very old con where you simply send every possible prediction to a different person, and progressively eliminate all the ones that got wrong info, until you have one left who has observed you make a series of impossible predictions.
There's a "card trick" magicians often like to do which is to simply ask the spectator to name a card at random. If they, by pure chance, happen to name the card that the performer already knows is on top of their deck of cards (eg because they peeked at it), then they reveal the card and it looks like they performed an impossible trick. If they named any other card, then the performer simply does another trick (perhaps looping back to the named card later or perhaps simply ignoring it or making a joke). Its a good opener: if it works, it grabs everyone's attention for the remainder of the act and if it fails, nobody will even notice as they get distracted by the actual routines.
Every so often, pure random chance means that they get it right and it looks incredible to the spectators who witness it.
Even better, do this with a card that has a higher likelihood of being chosen by the spectator—the ace of spades, for instance—and you'll see the trick work far more frequently than simple probability would suggest.
Oh wow, I'm glad I'm not the only person who noticed this. I was a Quora top writer early on and I found his work maddening. It was glib, persuasive, confident, and often incredibly wrong. The popularity of content like his was part of what made me eventually quit.
> If one were to run the numbers of just 2% of the population in China owning some fraction of Bitcoin, this could raise the single Bitcoin price to above $10,000.
I mean ... faulty reasoning or not: Bitcoin is currently at $10,240.
Of course I guessed he didn't believe his own reasoning, as when the article was written in November 2013 Bitcoin was $600 and he could have tenfold and more his own money.
It's also been at $20k. You could basically throw a dart at a wall and hit a price that bitcoin has been at some point. Just a coincidence you read this today.
It might be a coincidence, but it's not the kind of coincidence where you can give a number that sounds like a legit prediction but is sure to be correct at some point in time. Which you seem to have implied in your post.
Had he given a number above $20k he would't have been right at any point in time. And bitcoin might have never hit $10k, which would also left him being wrong.
Which is less than half of the peak value, I don't know what your point is? If we were having this conversation a year ago or a year from today, his guess would have been way off. That's why its a coincidence.
What's humorous to an old hacker like me is that these guys live so far inside the sandbox that the only thing they can write about is something like Square (that's some D2A credit card reader that plugs into a jack on old smartphones that have headphone jacks, right? --what "magic") With his UWB aspirations, he is at least aiming in the right direction! His little blurb is enough to make the physicist in us cringe but it's definitely good enough for business people that might hire him to use Apple and Google APIs.
The funny thing is almost everyone I've walked through using find my iPhone over the past few years assumed Apple already had this capability and ended up disappointed if they were looking for a device that was not sitting out in plain sight.
The noise originally came out of the currently-active output device, which in my case at least once was AirPlay, not the iPhone itself. I believe they eventually fixed that bug.
Probably because the Apple products to make use of it aren’t ready yet.
There are references in OS betas to assisted reality head-sets and a Tile competitor, which will both likely access this chip- but since they aren’t ready yet the demos for them were spiked and other references to the underlying tech scrubbed.
I'm guessing they won't release them until they have a new iPhone to try to sell.
It wouldn't really work if the new products only work with a product they released 6 months ago - there's little incentive not to just wait for the new phone anyway (Unless they expect everyone to be upgrading?).
No, because by the time they release anything the iPhone that was just released will be at least half a year old.
This might be a feature worth upgrading over, but who wants to upgrade in the middle of a cycle? The only people who would benefit are those who already have this new iPhone anyway.
So with all of these amazing attributes, why did Apple not announce the Apple A1 Chip? I assert it is a confluence of things:
The iPhoneOS software needed is not yet released
Apple will release AppleLocate tags for holiday shopping 2019
Apple had too many things to announce at this Apple Event and this would take too much time
Apple is aware of the privacy implication many will cast and wants to spend more time to explain
Other issues I can not present at this moment in time
Excellent writeup. The post mostly talks about navigation in indoor spaces and U1-to-U1 communication, but I hope it fixes Google/Apple Maps not knowing which direction I'm pointing my phone and constantly re-updating.
Apple Maps already have a mode which knows where you're pointing your phone (based on built-in compass). As far as I can tell, Google does not, and it figures it out when you move.
That's odd; when I visited Japan this past spring I used Google Maps on my iPhone practically everywhere while walking, and the biggest helpful UI element was the blue compass arrow that would point wherever I was facing. It was generally quite accurate as I can't even remember it ever really pointing a different direction from where I was actually pointing.
I have it happen in downtown Chicago, although I think it is more that the positioning is not reliable enough that I’m sure which street I’m currently facing.
Yeah, for some reason the compass chips or Google Maps sucks so bad with orientation-showing. I suspect it's the latter, because Google Sky Map works great with my phone.
>The new Apple‑designed U1 chip uses Ultra Wideband technology for spatial awareness — allowing iPhone 11 to precisely locate other U1‑equipped Apple devices. Think GPS at the scale of your living room. [...]
Think massive geopositioning satellites orbiting my living room and communicating with my iPhone? Neat!
I don't know why people are downvoting you, there's surely a privacy implication with such tech. I'm surprised there aren't SDR-based occupancy detectors sold to thieves, that say if people are home or not based on BT/WiFi and in future U1 UWB.
I hope this will become phenominally cheap, as I would love to see this thing used more in stores.
Here's a pie-in-the-sky example: I almost always buy clothes in-person. I want to see what I'm getting, and want to avoid the issues (discussed in other, mostly Amazon-related, posts) that come from buying stuff online. But I often run into trouble finding the specific item (make, style, and size) on the shelf. If each item were tagged (possibly as part of the anti-theft tag), I could be led directly to where the item is.
Here's a more realistic example: I want to buy an DB9 null-modem adapter from Micro Center. I can find out in general where it is. But if a tag were attached to the shelf, I could be led directly to where the item should be.
Of course, there are tons of issues (among which cost is but one). But it's nice to know that things like this may be possible!
>Here's a pie-in-the-sky example: I almost always buy clothes in-person. I want to see what I'm getting, and want to avoid the issues (discussed in other, mostly Amazon-related, posts) that come from buying stuff online. But I often run into trouble finding the specific item (make, style, and size) on the shelf. If each item were tagged (possibly as part of the anti-theft tag), I could be led directly to where the item is.
Cool idea but tricky execution. There'd need to be a way to "claim" that you already picked it up (maybe NFC from phone to specific item you pick up?) otherwise you get people chasing each other throughout the store.
I'd figure it'd work better just to embed one in the price tag for the item on the shelf or rack it's located on; at that point it's a matter of the store being kept as organized as possible so stuff doesn't get mixed into other areas.
Just an observation: half a century ago or so, things like that were possible; those shops would have employed people who could lead you to where the item was.
Said person would even have directed you to a competitor’s shop if they didn’t sell the article you’re looking for, and might even have done that if they did sell it, but didn’t have it in stock.
Hmm, that happens to me when I visit Home Depot. Employed people will lead me to the item.
Interestingly I also use Home Depot's online store just to look up the aisle/section/shelf. Stores can have a locator system that doesn't need active electronics.
It's really hit or miss in my experience. I've had several occasions at both Home Depot and Lowes where I ask an employee for help and the response is "We don't sell that." 5 minutes later after walking the aisles, I find it.
One example, I had recently rented a townhouse with very overgrown brush in the back yard, thick as a jungle. My weedwacker wouldn't scratch it, so I wanted a machete. The home depot employee acted like he'd never heard the term "machete" in his life. I explained what it was, and he shrugged, muttered something about big knives being illegal (untrue and ridiculous), and walked away. I found them in the gardening section.
I've also had some very good experiences. One immediately identified a thing I needed but had no idea of (a coupler to join a PVC waste pipe to an iron one). Another recently spent 20 minutes trying to locate a thing that inventory said was in stock but wasn't in the assigned bin. (It's not his fault that the system can't distinguish between items on display and items stuck, unsorted, in the high shelves requiring a front-end loader to reach.)
It's very hit and miss. The biggest issue is that it can be very hard to locate anybody at all.
Considering how hard it can be to find someone at Home Depot(who isn't occupied with someone else), I think I'd rather find the items myself. I'd prefer to have fewer expert level employees who I can consult with than more human-directory but otherwise useless to me employees.
Is this level of precision actually necessary in a well-maintained retail store?
I go to and use the Home Depot app in part because one can search for the aisle and section that any item is in for their particular store, and quickly find what they're looking for. The store near me is kept very well maintained, and the combination of aisle and section has never been wrong.
Attaching a radio beacon to a storage container or duct tape would be interesting, but it's just as useful to know that it should be in aisle 7 section 21, where I can go ahead and find it.
This is a problem that retail businesses could solve, but they don't want to. Why is the milk in the back of the store? So you spend more time there. The more time you spend in the store, the more likely you are to purchase more things.
Stores are often intentionally designed in a way specifically to stop direct access to the kinda of thing you'll just "drop in" to pick up. They want you walking past and seeing all the other things you could buy.
They want you looking around forever and hopefully buying some other things you didn't come in for specifically.
Same reason (I've heard) that milk is in the back of the grocery store. They want you to walk past everything and maybe notice something on the way to pick up that quick gallon of milk.
Throw one of these into the Apple Watch (or a gamepad) and you have something that works like the Oculus Touch Controller. Put one in your AirPods and you have head tracking.
Could be a powerful addition to their future AR hardware.
The biggest problem is finding Airpods that are in the case. In the case, they can't connect to a nearby phone. Out of the case, you can get in range, connect, then play a sound. I have lost my Airpods in the case a few times, and never lost a single Airpod out of the case. I hope this solves the lost-in-the-case problem.
That's Apple's Tile, actually. Having something like this on a Phone may and will bring up cool ideas. I see the random games easily but just give it some time for it to be the foundation for a location intelligence startup.
How well does UWB penetrate a typical office wall?
At a previous job we looked into locating. If you had two items parked on opposite sides of a common wall, you couldn't tell which rooms they were in. Which was critical for us since we did NOT want people wasting time running into the wrong room to grab it (think automated defibrillators)
It is tuneable and at low bandwidth it can have a range of several hundred metres, maybe 100m indoors. The location element uses time of flight which is less effected by walls than signal strength methods so should be accurate to within a meter even across a large building.
That was roughly the accuracy of the WiFi triangulation systems we looked at. And it wasn't accurate enough. The typical US office wall with metal framing studs is not quite 5" (127mm) thick, which would have been our worst-case (with the equipment touching the walls on both sides).
When I played around with the Decawave chips mentioned in the article, not well. When going through walls there'd be some weird attenuation and maybe some multipath stuff that gave unusual results. Going through a wall could throw the results off by 10cm which I suppose would prevent you from knowing which side of the wall something is on.
I'm looking forward to the tags. I hope they make something super small that can be snapped into a range of accessories such as dog collars or suitcase name tags (some first party, no doubt a ton third party to follow including some wonderful and crazy things like a message in a bottle). Once other peoples phones can find my things reliably, a lot of anxiety I have about things I tend to loose will go away (for a small price).
(I used to love Tile, but as a product we hit its technological peak several years ago and the network effect does not seem to improving this incrementally as much as it used to)
I don't get the dog collar application, unless the dog is always indoors or somehow needs to trigger proximity based automations (feed me, go outside where I can't be tracked without equipping the fenced-in yard with anchor nodes, etc.)
Bono won't care as long as they don't pitch him to invest in it (like webOS)...also it has to play "All That You Can't Leave Behind" whenever you activate tracking mode.
My main use case for finding things only needs it to show what building and room it’s in. Did I leave it at the office, or in my car? If it’s in my office, it can only be in a few places.
How common are the use cases where you need to know it’s 56 feet ahead?
Perhaps I think I mainly need coarse location because I only ever invoke "find my computer" after looking for quite a while. If I used it every time I needed to grab my laptop, I might care more about precise location.
This article is interesting and informative, but feels oddly synthetic. It follows the classic Quora spill-the-beans-and-predict-the-future format but has some interesting quirks.
The "How does UWB Work?" section has paragraphs lifted straight from the patent applications. Some sentences make no sense, ex: "the U1 chip along with the results of FaceID/TouchID system is stored in the Secure Enclave"; the typo at the end calling it 'A1 chip' before going into 'holographic crystal memory', plus the final sentence: "We will once again leave the Mainframe computer and become cloudless".
I'd guess it is an elaborate promotion piece for the author and his companies. Nothing is free on the internet anymore.
No, there is a difference between people that genuinely want to provide good information, expand knowledge, etc and people that are just using hype buzzwords to generate sales leads.
Promoted Content is an industry term, not just a phrase. Publishers will subsidize themselves by accepting money from companies to write promotional articles, even without disclosing it. There's a difference between that and an author just writing an article because it's their day job.
"The Decawave DW1000 Radio IC [3] for example, can move 6.8Mbps of data with an accuracy that is 100x better than WiFi or Bluetooth. It can reach 290 meters of distance with a very minimal power requirement with a 50x faster speed compared to standard GPS latency."
None of this makes sense as-written. It is either techno-blither or the author doesn't sufficiently explain "distance", "power", "speed", "better" or "accuracy".
The quote makes sense. He's comparing the ranging accuracy of UWB to other wireless protocols while also pointing out it's relatively high throughput and low power requirements. He's also comparing the latency of acquiring ranging information to that of GPS.
That mostly makes sense to me, it's a chip that allows for both spatially locating other chips and communicating with them, supposedly with more bandwidth and further range than WiFi or Bluetooth, while requiring lower power: https://www.decawave.com/dw1000/productbrief/
I agree with the gp, the quote makes little sense. It seems to take a bunch of facts from the product marketing but words them in such a way that is highly tortured and just doesn't make sense. Apparently you didn't parse it correctly either: More bandwidth? Even 802.11b had greater bandwidth. 6.8 Mbps is not more bandwidth than WiFi. What the quote is most likely trying to convey is that the radio has much better positional accuracy, which you didn't seem to glean in your paraphrase.
100x accuracy seems to imply that the data transmission is 100x more accurate which is not the cae. The product has high positioning accuracy. "can move 6.8Mbps of data with an accuracy that is 100x better than WiFi or Bluetooth." is a very bizarre way of saying that to be charitable.
"It can reach 290 meters of distance with a very minimal power requirement"
Very minimal? What is "very". Also in practice, the distance achieved is going to be dependent on power. So you can't get these maximum distances with minimal power. Also a bit misleading, as for non-positioning Bluetooth LE is superior.
"with a 50x faster speed compared to standard GPS latency."
This is a tortured way of saying that position acquisition is 50 times faster than GPS.
In practice, the laws of physics also apply. You're not getting 290m of 6.8 Mbps through any walls with minimal power.
Links to spec sheets always welcome. Thank you! Per the article's apparent claim: 6.8mbps is not a higher bitrate than WiFi. Or even close. 802.11ax (aka WiFi 6) offers three orders of magnitude greater goodput. I don't know what is meant by "100x better accuracy" - perhaps the author is making claims about the ability to locate a device with more positional accuracy than 802.11 triangulation via RSSI? Or 802.11mc? Or 802.11az? Who knows what the author meant?
Decawave UWB is really good for positioning accuracy--like sub-centimeter level accuracy. I've seen them scatter a handful of stuff on a desk and everything figures out where it is.
Yes, it's power consumption is quite good--compared with GPS and WiFi which have abysmal power numbers for small battery powered devices.
Compared with BLE--UWB doesn't look so great in terms of power consumption. So, most things only turn UWB on when they are trying to get accurate position measurements.
That sounds overly optimistic. From what I know, 15-20cm is the best accuracy you can achieve — which is still fantastic, mind you.
There is a lot of confusion around UWB positioning. Some articles describe TDoA (time difference of arrival) systems, which can indeed be very precise and the device being positioned can be low-power, but these require a nwetwork of beacons with precisely synchronized clocks, which is very difficult to achieve in practice.
Also, the DW1000 chip is a power hog. Yes, if you compare it to a GPS receiver it is still fantastic, but in low-power devices it is not easy to manage. As an example problem, you can't power a DW1000 from a coin battery: the internal resistance of the battery is too large, and it can't deliver the peak current that the chip needs.
What of the OEM specs of 10cm accuracy, only 31mA during transmission, 64mA in receive --which is usually not battery powered-- and its claims of months-to-years battery life using a coin cell? Did you overlook this or are you testifying anecdotally oR about your direct experience?
31mA is significantly more than most coin cells are designed to supply - while a new one may be able to reach peak currents of 50-100mA this significantly reduces battery capacity, and is only possible with a new battery (once it's partially discharged it cannot sustain these current levels any more).
I designed and built devices based on the DW1000, so I'm not "testifying anecdotally". If you have specific questions, please ask, but I won't be baited with generalities.
The author was on a podcast (Vector) talking about Siri and voice assistants, and he said multiple times explicitly that he would only elaborate on a particular topic to people who hire him as a consultant.
I have never heard the podcast but have heard the iMore guy who runs it. I'd look at the audience before condemning a consultant for promoting himself. For media outlets that focus on business people trying to get a high-level (Apple "G"genius bar) grasp on technology, I think it's okay to self promote. Now if he was speaking at IEEE conference on engineering, there would only be peers present and he would probably want to open up a bit more.
They were founded 1971 and 1985, respectively. The issue isn't everything getting worse, if anything, it's a lot of bad stuff crowding out what good things we have.
I think it's viable to determine the level of knowledge an author has on a topic purely based on the number of times the word 'technology' is used. In this case - this is almost certainly written by someone with little to no actually knowledge of the topic.
Location via time of flight is very accurate in terms of distance but not angle. It will require at least two septated devices to accurately pinpoint a third unknown device. And if those two baseline devices are moving then the accuracy is reduced. Android has support for location via wifi time of flight (802.11mc) where the baseline devices are your existing wifi access points. It makes much more sense indoors however it has two disadvantages: it won't work outside wifi range and it is much more difficult to make low power wifi devices.
I wonder whether you can fix this by using the positional tracking they developed for Augmented Reality (ARkit). If your device is moving, even a little, and the thing you are looking for is stationary, it should be possible to figure out exactly where it is.
All the AR capabilities in the iPhone is just a test bed for the eventual AR wearable. AR on the phone is just a clunky way to use AR, by holding it up to your face with a camera running. The U1 chip brings AR closer to how it should “just work”
So I assume they can superimpose a blue dot on an AR camera view to show me where in the room the kids left the Apple TV remote (again).
I was telling my kids about this at the dinner table tonight and we were cracking up imagining our dog eating one of these trackers, and you open up the app and the blue dot is following the dog around. Why Milo, why?!
Also missed the Groucho Marx reference, "Inside of a dog it's too dark to read." Which is definitely true when you're depending on UWB to provide "light".
I read it, twice, and the distance between "UWB environment" and "dog stomach" was enough that my pitiful excuse for a brain didn't recognize it after 30 seconds, and required reading this comment to realize what it meant.
HN tends to assume a ton of charity in the comments section, "did you read..." style comments are highly discouraged
171 comments
[ 4.3 ms ] story [ 207 ms ] threadHe'd write as if he had some sort of inside information and was a serious authority on the subject.
Anytime he'd make a new post about us, it would make the rounds on internal lists. Usually to the tune of "Oh wow, another Roemmele..." and we'd all read on in amazement at how much he'd just plain made up.
As insiders, we knew his posts were pure speculation, and were rarely even close to the mark about strategy or the technical details of our products.
Since then I cringe a little anytime I see one of this posts show up up. He completely lacked credibility when talking about Square. So I'm not sure why it would be different with Apple.
(just saying hello)
Every so often, pure random chance means that they get it right and it looks incredible to the spectators who witness it.
I mean ... faulty reasoning or not: Bitcoin is currently at $10,240.
Of course I guessed he didn't believe his own reasoning, as when the article was written in November 2013 Bitcoin was $600 and he could have tenfold and more his own money.
Anyway, the point is, it is a coincidence, nothing more.
Had he given a number above $20k he would't have been right at any point in time. And bitcoin might have never hit $10k, which would also left him being wrong.
There are references in OS betas to assisted reality head-sets and a Tile competitor, which will both likely access this chip- but since they aren’t ready yet the demos for them were spiked and other references to the underlying tech scrubbed.
It wouldn't really work if the new products only work with a product they released 6 months ago - there's little incentive not to just wait for the new phone anyway (Unless they expect everyone to be upgrading?).
like the one announced in the referenced iPhone announcement event?
This might be a feature worth upgrading over, but who wants to upgrade in the middle of a cycle? The only people who would benefit are those who already have this new iPhone anyway.
> Why Apple Did Not Announce The U1 Chip?
So with all of these amazing attributes, why did Apple not announce the Apple A1 Chip? I assert it is a confluence of things:
On my iPhone X, Google Maps clearly shows the compass direction with a blue cone emitting from my location 'circle'.
Think massive geopositioning satellites orbiting my living room and communicating with my iPhone? Neat!
Here's a pie-in-the-sky example: I almost always buy clothes in-person. I want to see what I'm getting, and want to avoid the issues (discussed in other, mostly Amazon-related, posts) that come from buying stuff online. But I often run into trouble finding the specific item (make, style, and size) on the shelf. If each item were tagged (possibly as part of the anti-theft tag), I could be led directly to where the item is.
Here's a more realistic example: I want to buy an DB9 null-modem adapter from Micro Center. I can find out in general where it is. But if a tag were attached to the shelf, I could be led directly to where the item should be.
Of course, there are tons of issues (among which cost is but one). But it's nice to know that things like this may be possible!
Cool idea but tricky execution. There'd need to be a way to "claim" that you already picked it up (maybe NFC from phone to specific item you pick up?) otherwise you get people chasing each other throughout the store.
I'd figure it'd work better just to embed one in the price tag for the item on the shelf or rack it's located on; at that point it's a matter of the store being kept as organized as possible so stuff doesn't get mixed into other areas.
Said person would even have directed you to a competitor’s shop if they didn’t sell the article you’re looking for, and might even have done that if they did sell it, but didn’t have it in stock.
Interestingly I also use Home Depot's online store just to look up the aisle/section/shelf. Stores can have a locator system that doesn't need active electronics.
One example, I had recently rented a townhouse with very overgrown brush in the back yard, thick as a jungle. My weedwacker wouldn't scratch it, so I wanted a machete. The home depot employee acted like he'd never heard the term "machete" in his life. I explained what it was, and he shrugged, muttered something about big knives being illegal (untrue and ridiculous), and walked away. I found them in the gardening section.
If you still can't find it (it's not 100% accurate) then just show them the app, with a convenient product picture and name.
Has worked for me when buying slightly obscure items there.
It's very hit and miss. The biggest issue is that it can be very hard to locate anybody at all.
I go to and use the Home Depot app in part because one can search for the aisle and section that any item is in for their particular store, and quickly find what they're looking for. The store near me is kept very well maintained, and the combination of aisle and section has never been wrong.
Attaching a radio beacon to a storage container or duct tape would be interesting, but it's just as useful to know that it should be in aisle 7 section 21, where I can go ahead and find it.
They want you looking around forever and hopefully buying some other things you didn't come in for specifically.
Same reason (I've heard) that milk is in the back of the grocery store. They want you to walk past everything and maybe notice something on the way to pick up that quick gallon of milk.
Almost nobody wants to use it that way (tested, worked and failed). It remains an idea that sounds great behind desks.
[1] https://youtu.be/UVpJouUyLBM?t=332
So it’s like Batman Begins?
It was the reason why UWB PHY was dropped from bluetooth
I'd love to see the fidelity of this. It seems unbelievable...
Could be a powerful addition to their future AR hardware.
At a previous job we looked into locating. If you had two items parked on opposite sides of a common wall, you couldn't tell which rooms they were in. Which was critical for us since we did NOT want people wasting time running into the wrong room to grab it (think automated defibrillators)
(I used to love Tile, but as a product we hit its technological peak several years ago and the network effect does not seem to improving this incrementally as much as it used to)
No, but they'll give everyone a free one, and then the blowback will be terrible.
How common are the use cases where you need to know it’s 56 feet ahead?
The "How does UWB Work?" section has paragraphs lifted straight from the patent applications. Some sentences make no sense, ex: "the U1 chip along with the results of FaceID/TouchID system is stored in the Secure Enclave"; the typo at the end calling it 'A1 chip' before going into 'holographic crystal memory', plus the final sentence: "We will once again leave the Mainframe computer and become cloudless".
I'd guess it is an elaborate promotion piece for the author and his companies. Nothing is free on the internet anymore.
None of this makes sense as-written. It is either techno-blither or the author doesn't sufficiently explain "distance", "power", "speed", "better" or "accuracy".
100x accuracy seems to imply that the data transmission is 100x more accurate which is not the cae. The product has high positioning accuracy. "can move 6.8Mbps of data with an accuracy that is 100x better than WiFi or Bluetooth." is a very bizarre way of saying that to be charitable.
"It can reach 290 meters of distance with a very minimal power requirement" Very minimal? What is "very". Also in practice, the distance achieved is going to be dependent on power. So you can't get these maximum distances with minimal power. Also a bit misleading, as for non-positioning Bluetooth LE is superior.
"with a 50x faster speed compared to standard GPS latency." This is a tortured way of saying that position acquisition is 50 times faster than GPS.
In practice, the laws of physics also apply. You're not getting 290m of 6.8 Mbps through any walls with minimal power.
Yes, it's power consumption is quite good--compared with GPS and WiFi which have abysmal power numbers for small battery powered devices.
Compared with BLE--UWB doesn't look so great in terms of power consumption. So, most things only turn UWB on when they are trying to get accurate position measurements.
There is a lot of confusion around UWB positioning. Some articles describe TDoA (time difference of arrival) systems, which can indeed be very precise and the device being positioned can be low-power, but these require a nwetwork of beacons with precisely synchronized clocks, which is very difficult to achieve in practice.
Also, the DW1000 chip is a power hog. Yes, if you compare it to a GPS receiver it is still fantastic, but in low-power devices it is not easy to manage. As an example problem, you can't power a DW1000 from a coin battery: the internal resistance of the battery is too large, and it can't deliver the peak current that the chip needs.
That's a load of codswallop, mate. The following projects stand to prove otherwise:
https://www.gutenberg.org/
https://www.fsf.org/
Certainly understand your sentiment, though it's important to not spread silly FUD like "nothing is free, everyone's a shill, etc."
I think it's viable to determine the level of knowledge an author has on a topic purely based on the number of times the word 'technology' is used. In this case - this is almost certainly written by someone with little to no actually knowledge of the topic.
Not true with beamforming. Imagine the chip as a little rotating radar dish.
It was flakey when it first came out, mainly due to hardware compatibility, but it’s pretty robust these days IMHO
https://blog.exploratory.engineering/post/decawave/
The way we used it is a bit power hungry and doesn't scale well, but we worked out a design that would have.
I was telling my kids about this at the dinner table tonight and we were cracking up imagining our dog eating one of these trackers, and you open up the app and the blue dot is following the dog around. Why Milo, why?!
Did you read the article?
HN tends to assume a ton of charity in the comments section, "did you read..." style comments are highly discouraged