Genius or Just Plain Lazy? The Internet Coke Machine

Yesterday a discussion in a meeting reminded me of Carnegie Mellon University’s Internet Coke Machine. Histories of the Internet of Things (IOT) point to this machine as the first example of a networked, remotely-controllable device. These days we’re comfortable with the idea that we can, and do, turn house lights off and on via smart phone while we’re vacationing in Bermuda. Mr. Best in Show can control quite a lot of the stuff in our house this way. For us this technology is a convenience, not a necessity, but remote control capability does contribute to a homeowner’s peace of mind. And there are plenty of examples where remote monitoring and control is more critical.

But a Coke machine? Who would care? I was working at Carnegie Mellon in 1982, when Internet Coke Machine made its debut. When I heard the story of the reason behind its development, my first reaction was, essentially, how lazy can you get? Not
how cool is it that graduate students figured out the technology to make it work. I’m sure a lot of inventions are prompted by the inventor’s desire to do a job more efficiently, or with less manual labor. But then sometimes a bunch of smart people, by solving a problem that might seem trivial, produce an invention with an impact far beyond its intended application.

Those of you who aren’t familiar with the story of Internet Coke Machine might be wondering what laziness has to do with genius, or vice versa. Here’s a synopsis. A group of computer science graduate students got tired of walking one floor down from their offices to the nearest soft drink vending machine, only to find that their favorite soda was sold out or not yet chilled enough to drink. Wouldn’t it be great, they thought, if we could find out before we make that onerous journey that we’d be disappointed? Remember, this is 1982, the infancy of the internet. As I recall, Carnegie Mellon’s campus network was for some reason linked to the one at Columbia University. Remote access was via 1200 baud modem, using a dumb Z-19 terminal, and the campus network ran on linked DEC-20s. I know; I’m showing my age here, but I digress.

A group of four graduate students (Mike Kazar, Ivor Durham, John Zsarnay, and David Nichols) took on the challenge of figuring out how to avoid fruitless trips to the vending machine. Each of the four undertook different parts of the project. Zsarnay installed microsensors, each hooked up to one of the local Ethernet networks, to detect how many bottles of soft drinks were in each of the six columns. The rest of the team wrote the software that gathered information about the status of the machine’s contents. They were able not only to tell if a column was empty; they could also tell how many bottles were in a column and infer how cold the bottles in a column might be.

To check Internet Coke Machine’s status, all one had to do was to “finger” the user named “coke.” (Finger is a network command used to determine whether a user is logged in or not.) Ivor Durham is credited with modifying the finger protocol to make it deliver coke machine status information. A serious programming book, Expert Programming: Deep C Secrets, has a section on Internet Coke Machine.

The picture above is not the Coke machine. To see a picture, use this link.

This invention spurred the development of a number of other wired soft drink machines around the world. In 1991 Cambridge University’s computer laboratory staff set up what might be considered a technological improvement: a coffee pot with its own webcam, so caffeine addicts could monitor coffee levels. Sadly, this camera is offline. The lab auctioned off the last of the physical coffee pots, a Krups, to Spiegel Online for 3,350 pounds; it’s now in the German Museum of Technology. The whereabouts of the original Internet Coke Machine are unknown.

Out of curiosity I searched to see where the original four graduate students ended up.

• Mike Kazar is the co-creator of the Andrew File system, among other achievements. He received the 2013 IEEE Reynold B. Johnson Information Storage Systems award.
• Ivor Durham joined Adobe Systems in 1986 as Director of Engineering and went on to work with two other companies with similar businesses.
• David Nichols was working at Xerox PARC in 1990.
• John Zsarnay was with the Robotics Institute at Carnegie Mellon in 1990.

Are there other CR4 members who were at Carnegie Mellon in the era of Internet Coke Machine? Or did you attend or work at another location with one of these early IoT contraptions? Let us hear from you!

References

http://www.cs.cmu.edu/~coke/

http://knowyourmeme.com/memes/internet-coke-machine

Image credit:

Wikimedia Commons