Hacker News new | past | comments | ask | show | jobs | submit login
What is the new Apple U1 chip? (quora.com)
507 points by MaysonL 10 days ago | hide | past | web | favorite | 161 comments





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.


"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".


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).

wouldn't you pair it with an inertia detector so it only transmits a pulse (at 31mA) when it moves by a significant amount?

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 DW3000 is accurate to under 4cm -10cm actually., indoors in line of sight.... and works for years in a watch battery.

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.

Could it be generated with a GPT-2 style model?

[flagged]


Well I am very familiar with the decawave, and I agree with the GP. This quote is bizzaro, as I mention in detail in another comment.

> but feels oddly synthetic

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.


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.

Every article is a promotion for the author and company it’s written by. No need to point this out on every article.

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.

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.

Then point that out instead, exclusively

I believe that was the intent with "promotional piece".

> Nothing is free on the internet anymore

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."


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.

While some of the things the author says on quora may make sense, he is a charlatan. His name is Brian Rommele and he used to have 300K+ Twitter followers. When Twitter cracked down on fake followers last year (or the year before) his count dropped to 70k. His tweets like crazy and constantly retweets his own tweets and retweets those again. He gets 6-10 likes per tweet which is very strange considering the follower count he has. He's constantly saying he invented incredible technology but never shows it (or shows something that could be easily and likely simulated). I wouldn't trust him to pick up a subway sandwich for me. Sorry for the rant, but I think the internet is giving these types of people louder voices than they deserve.

This is one of my major gripes with quora, I feel like it attracts and rewards these types of folks who are then immediately trusted thanks to their very high “internet point” scores

Yeah, sites which permit upvoting of posts and comments and display a total of upvotes received on user profiles are really cancerous.

I wouldn’t go that far, but when vote count is the primary means end users rely on to determine whether or not someone is “correct” then yeah, it can be problematic. I feel the same way about SO, but I wouldn’t extend this criticism to reddit or hacker news.

It's still a compromise when you're depending on people to create user-generated content for you.

One could argue this article is a net-gain for the community as a whole, even if he is ultimately hawking buzzwords and his consultancy. We're still getting this long-form exposition with a bit of research into Apple patents and other direct sourcing from both Apple and others collected in one place for free.

Content marketing can provide a lot more value to the world than most forms of marketing.

Additionally, 99% of people are only going to read the first half where he explains it's an ultra-wideband chip with a few examples then move on. They aren't going to dig into his futuristic predictions and other consultancy nonsense, or even look at his name.

Although having to skim through crap like this in order to get to the meat is a bit annoying:

> Many folks in the payment industry including disruptive startups thought me insane and went about becoming redundant when Apple Pay was released. Of course I had far more basis than a single Phil image. History is about to repeat itself.

This is bad enough to make it almost unworthy of HN. But otherwise it did the job of answering the question well enough.


For more information about the author, their personal site: http://voicefirst.expert/

LetsEncrypt is so simple in 2019 it really makes me cringe to see sites served over HTTP, especially when the site belongs to some self proclaimed “visionary”

I wouldn't go so far as to condemn him simply for trying to get his message out with Twitter trickery. Twitter itself is a giant freak show with "soft bans" etc. and would have already gone the way of My_ if it hadn't been for a certain politician using it to bypass the talking heads on TV.

The elephant in the room is that UWB positioning is indeed very precise, but only if you have a number of well-positioned beacons with known precise locations working together. The article is all giggly about the rosy future, but having worked with UWB (specifically, the Decawave chips mentioned in the article) I would be much more restrained with predictions. Yes, this is really good technology, but nowhere nearly as easy to use as it seems.

I'd be very interested in technical details: are they using 802.15.4a? Do they use the spectrum above 6GHz?


I know very little about any of this stuff, but it occurred to me that Apple already has a few technologies in play by which devices can sense the layout of their surroundings: HomePod can determine its proximity to walls by listening to how sound reverberates in a space, and iPhones map depth by comparing data between by dual cameras as well as by dot projection. Could these technologies potentially be used to eliminate or reduce the need for multiple UWB devices?

You don't need to know the positions of the beacons unless you want an instant location fix. You can use SLAM with inertial odometry or visual inertial odometry to find a precise position using UWB beacons at unknown positions. SLAM algorithms will also give you the positions of the beacons. The catches are that it requires compute, and it requires you to move around for the model to converge.

Is it any good for the where-is-the-dog use case in the article, where you might want to find something at opposite side of the house, presumably without a network of beacons to help?

can the signal penetrate objects? for example, when a receiver is hidden underneath cloths or sealed in a bag?

Yes, to a certain extent. Frequencies above 6GHz are easily attenuated, especially by anything which contains water (so, human bodies), but bags and clothes are fine.

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?!


Inside of a dog, there's stuff that absorbs radio waves. It's not a good UWB environment.

UWB?

Ultra Wideband.

Did you read the article?


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


UWB was repeated 39 times in the article. At that point, the "did you read" question(with the assumption he didnt read) become legit

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".

or Conan's AirPods commercial https://www.youtube.com/watch?v=z_wImaGRkNY remember that "Quora" predicted Apple would be putting U1s in AirPods too

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.

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.


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.

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.

His post on Bitcoin from 2013 is pretty funny: https://www.quora.com/q/kyvdkhixnlhvdjdp/There-Are-1-3-Billi...

> 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.

Not if you hit $30k, $50k, $100k or any of the other prices bitcoin has never been traded for.

On my wall the numbers only go up to the actual values bitcoin has traded at. And $10k today is only half the peak.

Anyway, the point is, it is a coincidence, nothing more.


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.

.. not yet. I'm sure it'll hit all of those numbers for at least a few minutes in the future.

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.

what is kelp?

(just saying hello)


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".

It was Quora - are you sure they didn't put "expert"?

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!


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.


Look it up in their mobile app. Tells you the aisle it's in and how many are in stock.

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.


Try being less specific, like "I need garden tool for cutting plant.", sort of like how Target.com works.

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.

Pretty soon they'll have us checking-in at kiosks and then waiting at carousels for our items to (very, very slowly) come out.

>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.


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.

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.

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.


Couldn’t you just ask a salesperson where the item is?

I assume this would be awful for retail.

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.


Quite a lot of indoor localization/positioning attempts rested their hope on such use cases.

Almost nobody wants to use it that way (tested, worked and failed). It remains an idea that sounds great behind desks.


Locating your Uber would be a great use case.

It may well become phenominally cheap... for Apple. Margins, baby, margins!

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.


My wife is constantly hunting for one of her Airpods. Just having accurate location tracking system for them would be amazing.

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.

It says it's 50x faster than GPS when it comes to latency but that still is too slow to be used as any sort of controller.

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.

> It will require at least two septated devices to accurately pinpoint a third unknown device.

Not true with beamforming. Imagine the chip as a little rotating radar dish.


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.

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.)

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”

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.

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.

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.

Whether a maps app has compass capability enabled depends on the parent hardware: in particular, if the hardware contains a magnetometer.

On my iPhone X, Google Maps clearly shows the compass direction with a blue cone emitting from my location 'circle'.


All iPhones since iPhone 3G (IIRC) have a magnetometer.

On Android, Google Maps can orient itself using the camera. https://mashable.com/article/google-maps-augmented-reality-w...

>It is entirely possible to build a useful AR/MR/VR map of any indoor space using the Apple U1 chip in just a few minutes few minutes.

I'd love to see the fidelity of this. It seems unbelievable...


I've always dreamed of a magic wand device that I could point at things and give commands. For example, point it at the window and say "close blinds". You could even add gestures to do it silently. With this UWB tech, accelerometers, a mic, and voice recognition, it seems like a viable product. Of course, I think Apple intends the phone to be the magic wand.

Huh, that's something actually interesting about the new phones. I wonder why it wasn't talked about at all on stage?

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?).


> they won't release them until they have a new iPhone to try to sell.

like the one announced in the referenced iPhone announcement event?


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.


The author asked the same question:

> 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:

    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

UWB is being touted as the best solution for next generation car keys, as all wireless solutions for car keys to date are susceptible to relay attacks (exception being car is online and controlled by an app). With an UWV chip in there, Apple could store your car key credentials and use their phones as a secure key.

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)


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.

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).

If it didn't penetrate you would have been better off!

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.

just make it sound the noise :P

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.

MKBHD at 5:32 talks about this after the event [1]. He talks about directional airdrop sharing features that the U1 allows.

[1] https://youtu.be/UVpJouUyLBM?t=332


More likely:

“Hey Siri, we lost Spot the dog, do you know where he is?”

Siri: "I found this on the web for 'we lost Spot the dog, do you know where he is'" displays google results

Seriously Apple, Siri is awful in its current form.


I once asked it to convert a currency amount ("Hey Siri, how much is $458 US in Canadian Dollars?") and it just gave me the wiki for the US dollar. You'd think currency exchange would have been one of the cases they would have tested.

This chip adds support for UWB to iPhones. Apple's new U1 chip uses UWB technology for spatial orientation, allowing the iPhone 11 to pinpoint the location of other Apple devices equipped with U1. Imagine GPS on the scale of your living room. Therefore, if you want to share a file with AirDrop with someone, just point your iPhone to this device and it will be the first in the list.

This just reconfirms my view that nobody can really touch Apple when it comes to tight software and hardware integration.

On the hardware front, very few companies have the capability to design and manufacture their own chips. Apple does it repeatedly.

The article also points to the very long timeline they can adopt in development. Some of these concepts have been in play since the early 2000s.


I copied this amazing comment from a HN user a few days ago.

I'll shamelessly paste it here since it's relevant to your comment:

Lots of people disappointed in the new iPhone. The reason I see is simple; Apple has long been outsourced a large fraction of its hardware innovation capabilities to other companies rather than having a full vertical ownership of the production line, unlike its competitors (Samsung, Huawei, etc). This works very well when most of the required technologies are already there for bringing their idea to the reality so Apple doesn't have to push the state of the art for the manufacturing technologies. Multi-touch, Retina Display, Apple designed SoC were all good examples where this strategy worked out very well.

The trouble is that now most of the low hanging fruits are gone and the rest of innovation opportunities lie within the manufacturer side and require non-trivial investments. For instance, getting rid of notch requires camera under screen technology. This is being developed by Samsung, their competitor. The same thing applies to fingerprint sensor under screen. While all the competitors are shipping 5G in their flagships, iPhone 11 couldn't ship 5G due to their hard dependency on Qualcomm. In short, the current landscape doesn't allow Apple to keep itself on the bleeding edge in the smartphone business.

I'm curious about how Apple will address this problem. Disappointingly, I haven't seen any positive signal to indicate that Apple has a good plan to address this issue. It first tried a high-price, even-more-premium strategy and this turned out to be a disastrous one. Apple now tries to expand into the services business and chooses to be a competitor to its own ecosystem by exercising its dominant position. I'm pretty sure that this plan will work very well, maybe too well sufficient to de-prioritize the iPhone business just enough to keep its marketshare around 3~40% and make no more commitments. I hope I'm wrong.


For those interested in using one of the DW1000 chips in their own projects, there is a hackaday project[0] for an Arduino compatible indoor navigation system.

[0] https://hackaday.io/project/7183/


There is one major big patent troll entity that goes around and sues anything UWB.

It was the reason why UWB PHY was dropped from bluetooth


I suspect Apple has the resources to fight simple trolls?

Broadcom?

Washington Research Foundation and another party named by some guy's surname and word "interests" in it. The later was made of 3 random troll lawyers.

How come we’re only starting to hear about UWB now? Has some development made it more feasible? As in why haven’t companies used it in previous products where location tracking has been an issue

> The new Apple‑designed U1 chip uses Ultra Wideband technology for spatial awareness — allowing iPhone 11 to precisely locate other U1‑equipped Apple devices

Interesting. Will this thing be always on and broadcasting my phone's precise location? Seems like a rather privacy-sensitive feature.


Apple has talked quite a bit about privacy when tracking devices, in preparation for iOS 13.

https://www.wired.com/story/apple-find-my-cryptography-bluet...


I’m trying to imagine a world where this technology is ubiquitous, but I can’t shake the feeling that it’ll be easily exploitable from the perspective of a nefarious actor (e.g. a jamming device of some sort). I just can’t see safety-critical systems coming to rely on it

or used by apps to track more people and things you thought were untrackable, all silently allowed by apple APIs, like iBeacons, Deep Links and more.

sigh.


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?


Some of us are a little less organized than others. Knowing which room my keys are in is remarkably unhelpful.

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.

I spent a solid half hour looking for my wallet other day. It was on a window sill hidden by the blinds which had been lowered.

Perhaps a future iteration of the Apple credit card will feature a U(x) chip, automatically embedding a beacon in your wallet.

Wonder if they will run into trademark issues when they move on to the next version of the chip.

> Wonder if they will run into trademark issues when they move on to the next version of the chip

No, but they'll give everyone a free one, and then the blowback will be terrible.


I doubt Lockheed cares much about an obsolete spy plane that was secret for most of it's existence.

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.

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.

We tried the DW chip last year. It works astonishingly well.

What was your application if you can tell us?

Indoor positioning. Just a prototype. We wrote a blog posting about it :

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.


This article deserves a better place than a Quora answer :) Why not a blog post of a proper article in MIT or somewhere similar?

I misread it as UI chip, as in it handles the User Interface and frees up the CPU for other things. But it is the U1 chip instead.

I see how it measures distance and direction, but how does it estimate height? Are there multiple antennas in the phone?

Dang combine 5G with this and the all seeing eye can follow you around a room with some serious precision.

>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.

Bought from Samsung ;)

time to figure out how to jam these signals. I don't want my 3d location tracked by every guided missile out there

The Biggest Apple Announcement Today Was What Apple Actually Didn’t Announce—Yet. HaHA!

I wonder when they’ll get around to fixing airdrop? Nearby People just don’t show up most of the time.

Likely related to your or their privacy settings.

It was flakey when it first came out, mainly due to hardware compatibility, but it’s pretty robust these days IMHO


Read in the voice of Michael Bolton from Office Space:

So it’s like Batman Begins?




Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: