We’ve already asked this question many times. What is innovation? What does it mean to build/sell/buy something “new”. And inevitably, as we ask this question, we leave a door open to interpretation. “innovation is in the eye of the beholder” I sometimes add. What seems obvious with consumer products however, is also true of technological products for businesses ; sometimes, “new” means “only better/faster” and it doesn’t have to be bad … this should make you think next time you shrug you shoulders while hearing “incremental innovation”.
What prompted this blog post is a piece found at datacenterknowledge.com which describes the data center of tomorrow. Or rather, it was some of the comments underneath (sometimes rather harsh) about whether that was or wasn’t new. The bone of contention was the following: while the author contends that future datacenters will be fully automated, the illustration of the Google data center he chose was dismissed by one of the readers as not being that new. True enough, I delved into Youtube and found quite a few old videos describing fully automated storage robots like this one:
And fully automated data centers aren’t to be seen in the future, they are already up and running as in Amazon glacier’s example. In this instance, backup and retrieval is performed by Amazon using a robotic tape library: “when you make a retrieval request, a robotic arm grabs the tape with your data in, slots the tape into a drive, and then your data will be transferred to a hard drive ready for you to access”. All is done in a 3-5 hour window and the principle is that you pay for data retrieval, while data storage is dirt cheap.
Yet, what Bill Kleyman describes is something entirely different. Instead of small robotised data storage room, he believes that whole data centers could be robotised on a massive scale, therefore making it possible for vertical as opposed to horizontal expansion. This is a new revolution I believe. Well… maybe. I first visited Whirlpool’s washing machine automated vertical storage warehouse in … 1986! Robots were moving up and down the alleys at breakneck speed and were able to store products and parts anywhere and very fast indeed. Whether you can apply this to a data center makes no doubt to me, and is certainly a step forward in better and faster data center management. Once again, innovation isn’t always about disruption, it is often about making things better.
innovation in the data center: how robotics is changing the game