Popcorn Indiana, the same company that manufactures the bags of kettle corn you might find in a convenience store, posted a project on their website called The Popinator. It’s a device you fill with popcorn, turn on, and responds to the word, ‘pop’ by firing a piece of popcorn into your mouth. Details on this build are scant, most likely because The Popinator doesn’t operate exactly as described in the video demo for the Popinator project.
We’re going to call this build a figment of the imagination of one of Popcorn Indiana marketing drones, but the idea behind the project is actually fairly interesting. The idea of using voice recognition to determine when the word ‘pop’ is said to turn on the machine is very cool. Using time of flight to calculate where someone’s head is puts this build into a category of awesome we’ve rarely seen before.
Despite all that coolness, we can’t help but think this project is simply an attempt to “go viral” and get a ton of publicity from random tech blogs using only a video camera and a few hours in Final Cut Pro. You’re welcome, Popcorn Indiana.
Surely there will be a ton of comments for this post arguing the merits of this build. You can check out those comments after the break, along with the official Popinator video.
28 thoughts on “Popinator Fires Popcorn Into Your Mouth, Is Probably A PR Stunt”
No “probably” about it. However, the gauntlet has been thrown down. Someone needs to build this.
More importantly, there should be a burst-fire mode.
Not a single mention on how the feeding system works
@0:53 you can see a black tube to suck in the popcorn.
@0:56 you can see the controll board controlling the servos (two stepper motor controllers)powered by an external supply (at about 1A, and 12v or 9v)
@1:04 you can see the board again uc chip looks like the same form factor as the ATMEGA 328 with an Ethernet breakout board (I see the plastic clip and the board has lots of pins) the other side of that eternet cable is another board seems it has a joystick(manual controll?).
the black wire is usb wich is powering that board and connected to the laptop.
the light board is connected by 5 wires.
these are only observations and assumptions by pausing the youtube video(720p) at key points and am in no way proving how real this device is just trying to state some observations to help.
I definitely don’t think that it’s using ethernet. It looks like they’re “using” an ethernet cable to connect to that game controller joystick in the background. I use scare quotes because I think it’s all fake.
They aren’t using the speed of light, but the speed of sound. This can be done in real time with analog VLSI.
Source: http://www.cs.indiana.edu/~jwmills/ANALOG.NOTEBOOK/klm/klm.html about 2/3 of the way down.
Trigger the localizer on the first P in Pop, assume everyone says pop at the same volume and use the signal to noise ratio to determine distance, then localize (again?) on the second P.
It seems plausible.
NOW PUBLISH YOUR SOURCE, COWARD POPCORN ENGINEERS!
I would say good PR; however it just made me think of the Cheetos “Break” bot commercial.
So using two microphones and determining the arrival time difference is a good start but that will only tell you the direction that the sound came from – not the distance it was away (or even the height it was in relation to the “Popinator”).
I think a device could be made that would do what they claim but I don’t see that it is this.
using asymmetrical microphones you can determine height by volume diffrence.
Elaborating on this: Imagine a device that would bake a piece of corn instantly and shoot it using its own power!
If you wanna go viral, at least do it in a sophisticated way…
From a marketing standpoint, this is pretty cheap.
Yup. Viral videos that aren’t humorous always leave a bad taste in my mouth. Am I supposed to be charmed or annoyed that I’m being bullshitted to over some Skymall-like concept? Give me my two minutes back :3
This reminds me of a student project of mine way back in 2004. The popcorn Fountain.
i don’t see how this thing would reload.
the firing seems plausable, but the device would have to use the volume of the voice to determine distance. wich seems kind of iffy.
I think for determining distance, you could have a camera look in the direction where the pop sound came from, and then look for the mouth. Everyone’s mouth is about the same size, so you’d be able to figure out pretty accurately, how far away they are.
or, even better.. find the head and then use a range finder
This is hilarious. It just needs a kinect and some high-powered launching system and you’ve got an even easier system.
For reloading a venturi suction tube only wide enough for a single piece of pop corn to fit through could be used. Burst mode is actually easier than single shot since you can just use a mechanism similar to that used by lottery drawing machines, or the ‘vortex’ or ‘cloud’ style bb guns. A cylindrical hopper with barrel and high volume air source entering/exiting tangentially creates a self agitating chamber and the popcorn eventually finds its way into the barrel.
The accuracy is definitely possible with just two microphones. I imagine it’s very processing intensive and requires a fairly complex model vs using more microphones, but given the distances involved, not improbable.
I disagree, I’ve done a lot of work with TDOA localization and I think this is well beyond the performance you can get out of two microphones.
The video would have been much funnier if every time someone said the “pop” part of “popcorn”, a kernel would hit them in the side of the face.
Haha, yeah that would happen a lot in the office of a popcorn company.
Anyone see the laser dot at the end of the video?
Fake but points go to at least making it appear plausible.
First Lake Woebegone, now Popcorn, Indiana. Yee Hawh! I think it’s north of Delphi, not near Val Hench for sure.
If you don’t have crop to process (drought) then invent wacky things to do. Did anyone buy popcorn futures?
Caution choking hazard!
nothing says fake like a hot piece of saturated fat shot at someone face
Two mics might tell you how far left or right to aim, but don’t tell you how high or low to aim and to what distance.
I think you need a third mic.
I’ll believe them.
I think you’re making this build out to be a lot more complex than it is.
If you assume a fixed distance, It’s all really pretty easy.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)