SnappCat - memed pics from your cat to your phone

The Concept

There is a weekly gathering on the left side of the mitten state (Michigan), in a small town called Holland. Three Electrical Engineers, sit around a table eating cheap cuisine, talking tech, and hacking away on the worlds next best cat gadget.

The device they are creating is compact, smooth, hip, portable, and contains a heart-warming touch that the user can specify during the time of purchase. It is powered with a small 5VDC supply; like the one that a Samsung phone uses. And while it is a piece of art in the physical, it is the software that adds a special touch to every feline friends home.

The device is Snappcat, and it is the only device that autonomously finds your cat, takes a photo, and delivers it to your phone with a fitting message.

This is not just another IP cam, or one of those other pet cameras that require people to spend time monitoring live video feeds while they are at work or the grocery store. Instead, SnappCat is autonomous, meaning all the user has to do is set it up, leave it alone; SnappCat will find the furry critters on its own.

We agree, controlling lasers remotely through an application which has a live feed to your pet is fun… for about a week. However, unlike other these laser gadgets, SnappCat doesn’t get old. It is always autonomously changing things up. The user will have a profile which allows them to personify their cat. This means the texts received can be mean, funny, outrageous, etc. Also the texts come randomly, just like another person would send them. Do not worry, the frequency of texts can be adjusted in the user’s profile, as well as the desired time of day.

Did we mention SnappCat doesn’t require the user to continually clean and fill a treat holder like other cat gadgets do? SnappCat is really the only “set it, forget it” monitoring solution available.

The Reality

While this whole SanppCat device sounds absolutely awesome, those three engineers that were mentioned above, aren’t quite there yet. Oh, and those three engineers have names and roles:

• William Gentry Glover - Electrical Engineer / CEO / Co-Founder
• Francisco Ceballos - Computer Engineer / CTO / Co – Founder
• Ben Petroelje - Electrical Engineer / CFO / Co - Founder

Since this publication is on the Hackaday page, we won’t repost all of our progress, but we will tell you where we are heading, and what we need.

Step One

It is our current goal to launch a beta/alpha-ish device on Kickstarter with an initial release of up to 100 units. We are planning small, because we are small. The belief is that this will let us scale up slowly and have a reliable product. By the way, that money Hackaday is giving away could really help us out with this stage. The quotes we have received on injection molding our current design are ~$10k. Also, the computer we would like to build for training those “autonomous” algorithms is ~$3k.

Step Two

After the initial release we would like to launch a large beta, and then launch one more time with the official product.  This is where SupplyFrame could really come in and help us learn what it really takes to turn SnappCat from beta, to a real product, and more importantly into a company. We desperately need coaching in business and marketing, and we believe SupplyFrame can do that.

Step Three

Well there really isn’t a step three in this very casual business plan, it just seemed fitting at the time of writing this. And while there is not a step three, and this business plan isn’t traditional, it does not include every nickel and dime, or an executive summary (we have that one tucked away in our file cabinet), it is an opportunity for us to tell you we are serious about SnappCat, and we really want this to succeed. We can humbly admit though, that we need Hackaday’s and SupplyFrame’s help, so please consider us in the Top 20 (preferably #1) for the 2017 Hackaday Best Product.

Hello all,

It has been a long time since I (William Glover) have updated anything on our Hackaday page, and for that I do apologize.

However, even though it has been a while, we have not stopped working! Since last time we posted:

• We have continued to update our image recognition algorithms
  • Let’s just say, I am becoming a believer that this is really going to happen. Every time we train, we are getting more accurate, and in turn getting less false positives. We are testing against random videos off media websites. It puts a smile on my face – it really does.
• Switched from the Pi Zero to the Pi 3
  • We are ramping up for a possible beta/alpha-ish release! In order to do this we needed to have hardware that is readily available. The Pi Zero W is just too difficult to get at this point in time.
• Designed a new case!
  • Yea that is right, the cases you all have loved, were in fact lovely. However, they did not accommodate the Pi 3, and l promise the new case will not disappoint! I’m not including any pics yet, because we are playing around with docks, and mechanical options which will be customizable by the consumer (which is you!!)… What is better than non-custom… Custom!
• Social Media
  • We are now on Instagram & Facebook – Make sure to follow us, please:
    • https://www.instagram.com/snappcat.official/
    • https://www.facebook.com/SnappCat/
• Got our first server up and running
  • We have acquired a long-term loaner server, and it is glorious. We are working on everything from how users interface with SnappCat, to how many users do we want to support (in the beginning); To infinity and beyond –Buzz Lightyear

Alright so my apologies on the lengthy blog, which contains no pictures… wait, I will add at least one picture. Anyways, this is what we have been up to, people of Hackaday; so love our page, follow us, comment, you know do everything awesome!

Meet The Team - Gentry (William), Ben, and Frankie.  Not pictured - Newton (cat), Zoro (cat)

It is the heart & sole to any good computer vision system... a good image database.

This is why our group has been working hard on expanding and building our image database. Stay tunned for more updates. :)

Over the past couple weeks the team has been working really hard to further improve our image recognition algorithms. This is a picture of Snappcat's latest algorithm at work. Thus far it seems to be really robust in medium to low light conditions, and so-so in locations with really intense light. We believe this is not so much a limitation to the algorithim itself, but more so because of the way the camera focuses in these conditions. More updates coming soon! :)

Changed the enclosure design, and had some fun printing on the Lulzbot mini; New purple filament :)

Started working on an enclosure. Here is the first revision render: