CandyCodes: Simple unique edible identifiers for authenticating pharmaceuticals

This kills the diabetic.

Though seriously, if you make the pills, you do not need to put anything on it, you can include colorized patterns in the pill itself during fabrication. Any mechanical problems are pretty much nonexistant. The pattern would also be volumetric and could be scanned after e.g. breaking a pill apart.

In any case - I feel increasing the availability of common medication is a much more useful goal. So a simple method to verify whether a counterfeit really does what it should do benefits humanity more than trying to beat a working counterfeit (if counterfeit is just code for 'unlicensed').

This is an interesting idea but it feels unlikely to be adopted -- every pill would look the same and branding/image (ethics be damned) are a huge part of pharmaceuticals.

Do you want generic Valium, or the Roche Valium that has the suave cut-out in the shape of a "V".

Xanax is a drug synonymous with "bar", the rectangular shape is locked with it's identity.

Somehow I don't see Big Pharma sweeping this idea up, however clever and useful it might be.

Good read though.

Related thread: https://news.ycombinator.com/item?id=31897530

How to further increase the costs of medicine... Issue should be addressed at the source, this idea sounds as good as arming teachers to make schools safer.

Still an interesting idea that could be used somewhere else.

> CandyCodes could play an important role in the fight against fraud in pharmaceuticals and many other products.

One wonders if these other products include recreational substances and black market branding, as it would be a fun project to write the openCV code to extract a Vornoi map of the image of the pill and determine if a hash of its matrix were found on something like public ledger. Perhaps using a Bloom filter. Phone cameras are good enough that you could use these maintain some integrity in a reliable black market brand.

I actually like this candy version better than other PUF proposals because a person can reason about it and prove it to themselves, whereas other opaque methods like crystals and chemistry reduce to "Trust us! I didn't quite catch your affiliation..." As a standalone product, I don't see this candy tech flying, but as a solution with open implementations, I could see it being useful.

This is cool, but I imagine those sprinkles might easily fall off or be shifted. The packaging they put edibles in are often crazy strong, but just the act of prying at the packaging or solving the puzzle might put your sprinkles at risk. Maybe temperature changes too would cause trouble.

Why not just print a key on the packaging? Or have each one wrapped individually in paper or what have you with a key. Might make better sense for other drugs. A database like this would provide some interesting info.

Does anyone else find it troubling that the simple act of scanning the pill would be a signal for who is consuming it? Even with reasonable amounts of PII removed from the stream to the validation service, they will know you have a cell phone and a general area where the scan is happening. It is more data and therefore, more marketing and very likely more price adjustments.

I think edible inkjet ink dots or differently dyed parts of the filler mixture (think he two colors of rice in some basmati rice from Indian restaurants) would be easier to implement and still keep the shape and overall color of the pill identifiable.

To add a different perspective about why someone might not know how legitimate the pills they’re buying are, here’s a photo of a Haitian street vendor with pills: https://i.pinimg.com/originals/e2/67/21/e2672163011a5853ae66...

The only real concern for people using medications is whether or not the label matches the molecule. The only real way to be sure of that is to have reputable testing labs capable of analyzing the substance and checking for additives.

As far as intellectual property in the pharmaceutical sector, those drugs were almost entirely created with taxpayer-financed basic research, and prices are typically inflated well beyond what a competitive market would result in, so chemically pure counterfeits, aka 'generic medication' will continue to have strong market demand.

If these are as durable as they claim to be, why not put the unique identifier on the sealed container instead of the pill?

Also mentionable that you can collect data on who took what and when.

These pills seem hard to swallow.

This seems strange at first glance, but let's entertain the idea for a second:

Here's the use case: You receive a package of pills / vials from a supplier (and have to pay $$$ for it) and you want to know if the vaccine / drugs are 'the real deal'.

This is the proposal:

By manufacturing the pills in a way that they have a physically complex structure that is hard to replicate, you can convert that structure into a code which you can then check against a database to see if it is valid.

Now, we compare that to just having, say, a printed code in the box that we check against the same database.

In the latter case, you can just open the box, photocopy the code, and make ten new boxes full of rocks with the photocopy in them. Each box will be 'valid' according to the database.

So, by making the code something you derive from the physical pill, you can avoid this problem.

...

> If enough multicolored particles (like these candy “nonpareils,” top) are randomly attached to objects, the odds of two objects having the same pattern of particles are essentially zero

Sure, if enough multicolored particles are randomly attached.

...but in those 1cm samples, I can count, just by hand, roughly what... like, 10 x 25 little color balls?

The problem of attaching 250 little colored balls in a specific configuration on a pill is simply 'hard'.

I could do it by hand.

So... that's not gonna stop anyone who really cares about doing it. Physically replicating the arrangement of a small number of highly uniform objects is a trivial task.

...but for a second consider the seemingly stupid idea of having a vial with a candy-code on the lid. The code is not linked to the vial, so there's nothing to prevent you tampering with the contents (eg. replace with saline) ...but, you can't easily just photocopy the lid.

You have to manufacture a new lid.

...

Which brings me to my 'huh'.

So, breaking it down to the most primitive form, we have an object which is a) difficult to replicate, b) easy to manufacture and c) can be used to derive a unique code.

There's... no reason this should be on pills. Forget that. Just ship it in the box.

Now you have:

- A box consisting of an object and a set of pills and vials

- An app that scans the object and produces a code.

- A database that links a code to a manifest of pills and vials

- The object cannot be duplicated physically.

That's actually a pretty interesting idea. It won't stop tampering, but it does allow you to assert single instance authenticity of the shipment.

You can take my box and create 20 that look just like it; you can replace the pills with rocks; but only one of them will have this magic object that says its authentic, effectively preventing you from selling the rest of them.

Maybe there's something interesting in there. I could imagine someone selling boxes of 'one-time-tokens' you could use this way... and all they would be is little blobs of plastic in some difficult to deliberately manufacture way.

...of course, this is based on the idea that it is difficult to recreate a physical object like this, which, with 3d scanners and printers, I'm pretty dubious of... you'd have to do something with your 3d structure to stop people just doing a surface scan... but then the object -> code mapping becomes problematic.

But hey, it's an interesting idea.

...except putting it on the pills. That's just daft.

It won't work. 1. Are they requiring users to check authenticity? 2. Have they taken into account stuff like 3d printers and fake verification software?

Interesting problem. I would have suggested to secure the bottles instead. Make a bottle that spits out a pill at an interval after authenticating the user (with one unauthenticated emergency pill per bottle) and tamper proof it by having a layer of air-pressure gap and conductive epoxy where if the pressure drops or voltage differential changes a solution rendering the pills useless is released(dilluted HCL?), that way you need to pierce it under a bottle specific air pressure an figure out the bottle specific votage for the epoxy. Sensors and all included I speculate under $30 shipped is doable for very high risk and addictive pills that cost a lot anyways.

I think the ability to detect that you are not eating a counterfeit isn't that important because if you were making counterfeits you would just put the taste on it as well.

however it might also help you to remember if you took the pill or not, as the taste might be another mnemonic helper for people who are perhaps temporarily overloaded and now can't remember if they took the pill or if they just thought about it and got distracted.

on edit: there is of course this -

>By taking a photograph of each CandyCoded pill after manufacture and recording the location and color of each nonpareil, a manufacturer can construct a database containing the CandyCodes of all known-authentic pills they produce. A consumer can then simply use a cellphone to photograph a pill and transfer its image to the manufacturer’s server, which determines whether the pill’s CandyCode matches a known-good CandyCode in their database (meaning that the pill is authentic) or does not have a match in the database (in which case the consumer is warned that the pill may be counterfeit and should not be consumed)

which seems like something I wouldn't expect the public to actually do, but maybe in an environment with many counterfeits incentives are changed.

on second edit: I guess the swift downvotes must mean I have now crossed a contingent of candy-tech true believers?