Pulpatronics tackles single-use electronics with paper RFID tags | Hacker News Reader

Pulpatronics tackles single-use electronics with paper RFID tags

dezeen.com

48 pointsalbertzeyer31 comments

> By contrast, PulpaTronics' alternative RFID design requires no other material than paper. The company simply uses a laser to mark a circuit onto its surface, with the laser settings tuned so as not to cut or burn the paper but to change its chemical composition to make it conductive.

Can RFID tags be lased into food and clothing, too?

"Laser-Induced Graphene by Multiple Lasing: Toward Electronics on Cloth, Paper, and Food" (2018) https://pubs.acs.org/doi/10.1021/acsnano.7b08539

Nicely spotted. I only managed to find more recent papers searching for "conductive carbon-based material".

PulpaTronics will be so cool if it works.

Even better would be to come up with some kind of combo RFID and barcode. And laser print that tag on the outside.

Straddling (or bridging) the RFID and barcode domains is a huge opportunity.

FWIU RFID cost has been a major limiting factor in supply chain traceability; pallet-level RFID is almost affordable?

> PulpaTronics estimates its tags will reduce carbon dioxide emissions by 70 per cent compared to standard RFID tags while halving the associated price for businesses.

From https://news.ycombinator.com/item?id=28691422#28701355 :

> There's probably already a good way to bridge between sub-SKU GS1 schema.org/identifier on barcodes and QR codes and with DIDs. For GS1, you must register a ~namespace prefix and then you can use the rest of the available address space within the barcode or QR code IIUC

https://schema.org/identifier rdfs:subPropertyOf^-1 duns, globalLocationNumber, gtin, isbn, orderNumber, productID, serialNumber, sku, taxID

GS1: https://en.wikipedia.org/wiki/GS1

List of GS1 country codes: https://en.wikipedia.org/wiki/List_of_GS1_country_codes

W3C DID: Decentralized Identifier: https://en.wikipedia.org/wiki/Decentralized_identifier

> FWIU RFID cost has been a major limiting factor in supply chain traceability; pallet-level RFID is almost affordable?

But I guess that depends on the type of RFID tag, isn't it?

For example, Decathlon already introduces tags with UUIDs with every single product they sell (even the cheapest socks), and they use them to inventory tracking and to scan your cart when you buy in store.

https://www.cisper.nl/en/case-studies/rfid-case-study-decath...

Of course, although business-wise practical, that can pose a security issue... You can track peoples clothes on the wild!

https://mastodon.social/@russss/102674489508219187

Is it the link between the Invoice and the ~serialNumber of the scannable-at-a-distance Product?

Most barcodes today only encode a ~sku (and not a serial number).

Minimally, to lookup additional product information from just an sku (similar to a shorturl app), consumers would need an API to retrieve and/or lookup from a conjunction of per-Organization named graph documents linking schema:Things' ~schema:identifiers with the RDF subject URIs of NutritionInformation records.

  - P: schema:nutrition 
    - d: MenuItem,Recipe 
    - r: NutritionInformation

  - C: NutritionInformation

https://schema.org/NutritionInformation

C: https://schema.org/Offer supersedes the GoodRelations vocabulary for Products and Services:

> For GTIN-related fields, see Check Digit calculator and validation guide from GS1.

I just checked one of the electronic invoices I had in my email, and for every line (be it a protein bar or a jacket) it includes an "RFID" field with what I guess it's the serial number contained in the tag. I don't have a reader for that frequency so I can't really confirm.

Edit:

An example for some chocolate cereal bars:

  Barrita cereales Chocolate x10
  Talla única
  Ref: 2610768
  RFID: 003869592404

That'd be for this product:

https://www.decathlon.es/es/p/barrita-de-cereales-clak-con-c...

Very interesting. Thank you for sharing.

I would have not thought per unit tracking was practical. The RFID wiki article agrees with my long held assumption of $0.05-0.25 per. (Caveat: My RFID knowledge is circa 2005, so very out of date.)

From a super quick skim reading, it looks like Decathlon uses Tageos. They come in a roll, so certainly could be a lot cheaper than $0.05.

https://www.cisper.nl/en/tageos-eos-300-impinj-m730-54x34-mm...

https://www.cisper.nl/datasheets/tageos/Tageos_Datasheet_EOS...

It appears Tageos is using aluminum foil as the conductor for RFID UHF. No surprise there.

So if PulpaTronics can ship non-metal RFID, I think they'd still be competitive.

Don't know exactly which exact model is decathlon using (once I get home I'll see if any of my equipment still has its tags attached), but I found tags at less than ten cents buying 5000 units rolls from a random US distributor [0]. So being French both Decathlon and Tageos, and with the volume they must buy, no doubt they can get to 5 cents or so.

Not using metal, I guess Pulpatronics could also do local factories more easily, needing "only" paper (and lasers). Easier to do JIT production or smaller batches I imagine.

  0: https://www.barcodefactory.com/tageos/rfid-labels/eos-241/o184c01x309003

FWIW Thermal labels last from 2-35 years depending on the thermal printer and the label.

There are thermal cable labels that wrap around e.g. ethernet cables

Can RFID tags be lased into food and clothing, too?

what about... people?

rfid tattoos?

maybe new kinds of handstamps to get into events?

(also I wonder what degrades the rfid circuits - wear or water?)

> Instead, the "chipless" PulpaTronics tag uses the geometric pattern of the circuit itself to convey the information. In the company's concept designs, for instance, it's a labyrinthine pattern of concentric circles.

So does this need special hardware to read? Seems like if so, it would be more expensive to install compared to traditional RFID systems.

> "It's basically storing the information in the antenna."

Since I'm pretty sure they can't use a laser to write transistors into paper, this probably means they just write different resonant circuits straight into the antenna structure. Their "concentric circles" are just different inductances and capacitances.

So I guess the reader externally excites the resonant circuit, and reads back the different frequencies at which it oscillates resonantly.

If this is the case, the reader would be significantly more expensive than current RFID tech - since current tech is using digital logic, and this is basically analog RFID. You basically need a vector network analyzer to read it out. Those are cheap-ish (below $100) if you stay at low frequency, but if you want to go above a GHz, things turn expensive quickly.

Also, while I can imagine how you could encode information into one tag and read it back out, I can't imagine how you would read back an entire basket full of tags at once (like you can with conventional RFID tags, which makes checkout at supermarkets ect. really convinient).

It seems like that's a different variant. "In addition to the paper circuitry, PulpaTronics also applied another of these experimental technologies...", so the "geometric pattern in the antenna" thing is a different tech from the "paper circuitry" thing.

EDIT: https://www.jamesdysonaward.org/en-US/2023/project/pulpatron... From this other article it sounds like there is no other variant, this is indeed just encoding in the antenna shape. In which case using the terms "RFID" and even "paper circuitry" are big stretches...

The geometric pattern appears to be a red herring. In the article one of the founders says "but it doesn't need to be scanned visually" (so it's not like a barcode).

In the Dyson page (nice find), the last slide shows the electrical response compared to a metal tag, using a vector network analyser, and includes the text "Over 1000 readings, our paper-based tag performed equally well as the copper tag".. which suggests it's intended to work with current readers.

The way I understood this is they developed the concentric ring pattern first, build a prototype using standard copper etching, and showed that it worked. Then they showed they could achieve similar results by fabricating the same pattern using a laser and paper.

Which means you need new readers. Readers that essentially are pretty fancy network analyzers.

If they want to use current readers, they need a tag with logic gates, one that can send a digital response.

My dream: to have every supermarket item embedded with an RFID tag so I can have a constantly up-to-date pantry inventory, with notifications of out-of-stock or near-expired products.

Yes, I am too lazy.

Would also be interesting for household garbage recycling at the recycling plant. If you can ask your garbage "what are you" and it responds "I'm a polystyrene yogurt cup with an aluminum lid", separating it gets really easy.

At least in some parts of Italy (I was in the Lombardy region) you pay an annual garbage tax where they provide you yearly with a predetermined amount of bags containing an RFID tag. You can still ask for more bags if needed, and I guess that in that case your tax would get adjusted depending on usage (which they track with the RFIDs). Kind of unsettling though that they could link your trash back to you.

There should be sufficient incentive for trash and recycling service providers to implement sensing and sorting now that "Making hydrogen from waste plastic could pay for itself" (2023) https://news.ycombinator.com/item?id=37886955

If manufacturers had to pay for waste management they would immediately reduce single use packaging consumption. Yogurt tubs instead of single serve cups. Beer bottles can be standardized, collected after use, sanitized, refilled, and put back onto grocery shelves.

In Germany they have to, but most of the plastic still gets burnt. Market failure. With paper, glass, aluminium and tin plate the numbers (and the economics) are better.

I'm surprised retailers like Whole Foods haven't added something like this to purchases made in store tied to an Amazon account. At the point of sale they have access to best-buy dates, etc (or could even add it during inventorying) and could use that data to alert you to restock with better granularity than sending a reminder based on your shopping trip frequency.

I wrote a system for this for my house.

Entry of the expiry date is manual, and I haven't implemented notifcations yet, but barcodes solve most of the rest of the issues.

Why isn't this possible with barcodes?

It is possible if the barcode would carry longer serial numbers for identification, not classification of the certain product. The dream of rfid proponents in the nineties was to push a whole pallet through a gate and get a super quick control of packing list at arrival. That would not be possible with optical scanning without unpacking the whole pallet. In production, under somewhat controlled conditions, Barcode ids for parts are used for decades.

Barcodes on products only have UPC/EAN information, which identifies a product, but not its expiration or batch.

Furthermore, I find that barcodes are a bit more picky w.r.t. how to scan them, while RFID is easier.

Depends of the orientation of the tag and the background interference from other sources (or other tags).

> This mechanism is similar to barcodes and QR codes in the sense that the information is encoded geometrically, but it doesn't need to be scanned visually.

They really need to give more details on this. Maybe there is some RF voodoo here, but I find it hard to believe it is practical. Also, how much data can be stored? And aren't most RFID tag programmable? If this requires a laser to be programmed, it might be a significant downside compare to the chip version.

Hard to find stats on how many are fixed vs programmable, but I suspect the majority are not programmable. Anything anyone calls a FOB is a fixed RFID tag. Many used for theft detection (library books, retail stores) are a fixed RFID tags. This seems to be tackling the Electronic Product Code (EPC)[0], which is popular in retail.

[0]: https://en.wikipedia.org/wiki/Electronic_Product_Code

Second that, it is way cheaper to keep the intelligent part in the database. With electronic keys it is convenient to frequently update the permissions of the bearer because it allows for offline locks, but for tracking goods, any serial number that is long enough is more than fine.

Reply on news.ycombinator.com