Ingredient List Scanning – A new idea?

Well it’s certainly been awhile since my last post announcing my intention to build the Healthizer app. I’ve been coding my little butt off for my startup company for several months, and as it turns out have had little time left in my schedule for anything else! However with me once I’m interested in something it never quite dies, it just goes on the back burner for a little while…

So now that our startup has published its app (I’ll post that link soon) I finally want to dig in and get on with this plan. Also I need a vehicle to mess around with some new development technologies I’m interested in.

Since I posted the app concept there have been quite a few apps cropping up containing similar functionality to the concept of Healthizer:

Don’t you hate it when you get an idea, and then find out someone else has already run out and done it way better than you ever could have?! Curse this manic business of high tech and all its brilliant people and wonderful apps! However, I believe there is still hope for the Healthizer.

  1. None of the existing apps use OCR technology (Optical Character Recognition – ie. computer reading) to parse ingredient lists. The preferred method out there for product lookup seems to be to scan UPC codes (ie. bar codes) and look up the associated ids in product databases. This method has a few drawbacks that I think would be avoided with my method:
    • it incurs a cost for accessing product databases through services such as Scandit.
    • a successful scan requires that the scanned product be registered in such a database and that the ingredient list in the database is accurate
    • it requires a working Internet connection, which has its own set of issues. For one, while the Internet is ubiquitous, having a good connection to it is still not a guarantee at all times (let me tell you), even in this day and age. Also, accessing the Internet in itself increases the cost of using the app, and slows it down.
  2. Most of these apps are not specifically geared at cancer prevention, with the exception of Think Dirty. Many of the apps are intended to assist shoppers make healthier decisions for weight-loss or lifestyle reasons, which means that the apps rate the healthiness of products based on some criteria. Therefore there is some subjectivity being introduced in their recommendations, and there is a question as to whether the user agrees with the criteria being applied – certainly not a straightforward question. Instead, I’d like to make an app that doesn’t recommend, it only points out to users the existence of a known carcinogen in the product they are considering buying, according to the product’s own manufacturer and lists of known chemicals published by mainstream authorities.
  3. Finally, it’s unclear whether the companies that produce these apps are for-profit companies. If so, there may be commercial motivations inherent in their recommendations. I feel that to gain users’ trust within the highly political fields of food manufacturing and cancer treatment, it’s important for the app to be strictly non-commercial, with a simple, straightforward algorithm, and transparent sources.

So I think there’s still room out there for a very simple, single-purpose app with a slick user experience that can be used very quickly and easily to flag a carcinogen-containing product with next to no fuss or fumbling or having to interpret the results. The main use case would be to scan products as you pick them off the shelf during your weekly shopping trip, so it has to be a streamlined experience.