Cameron Haight

A member of the Gartner Blog Network

Cameron Haight
Research VP
10 years at Gartner
30 years IT industry

Cameron Haight is a research vice president in Gartner Research. His primary research focus is on the management of server virtualization and emerging cloud computing environments. Included in this effort is… Read Full Bio

Coverage Areas:

IT Operations Futures: Will We Find An Aiko In Your Data Center?

by Cameron Haight  |  December 10, 2008  |  4 Comments

I just saw a recent news item on this.  I honestly don’t know if it is creepy … or fascinating (make sure to see the videos).  You be the judge.  While it did remind me of some past movies (see Austin Powers, etc.), it also got me to thinking which is always a dangerous thing.  If we can build an Aiko, then instead of just being someone’s um … friend, can we put robots like this to use in more critical roles?  Like … IT operations (sorry, but I don’t let my mind wander too far from my coverage area – wouldn’t be prudent).  It’s no secret that while some things are getting easier with the advent of SOA, virtualization and cloud computing, other things are getting more difficult.  We talk a lot these days about automation being key, but come on, compared to this, much of the automation technology I see evolving within enterprise management seems stuck somewhere between the stone and bronze ages.  The videos show the robot able to do a variety of tasks including amazing pattern recognition and rapid calculations.  So, if we’re trying to triage a particularly difficult virtualization problem, my bet is that Aiko could do this pretty quickly (at least going down all of the probable fault paths using something like an Ishikawa diagram).  I know many of us have often heard something along the lines of, “Well, if we can send someone to the moon, why can’t we (fill in the blank)?”  So my question is, “Well, if we can build an Aiko, why can’t we have truly far-reaching automation within IT?”  


Category: Uncategorized     Tags:

4 responses so far ↓

  • 1 Mike Malloy   December 10, 2008 at 7:30 pm

    That definitely is creepy. I think the limiting factor for automation has been the combination of complexity and volatility that IT operations face. Traffic, applications, databases, and a hundred other things change everyday. Teaching automated tools to assimilate and respond except in simple cases has simply been too difficult thus far.

    Now, I wonder if I can teach that sys admin to recognize my face…

  • 2 Cameron Haight   December 10, 2008 at 11:23 pm

    Thanks for your comments Mike. I agree that complexity is probably the major factor, but I would also state that automation (and associated analytics) are not areas that have generally seen huge investments by the major management vendors (I am of course not including acquisitions as this is not the same as innovation). Why? I’m not sure … it is expensive work to perform (often requiring lots of Ph.D.-level talent) and perhaps that also explains some of it. Additionally, it seems like there is still lots of other “low hanging fruit” to be gained with less cost or effort. I can’t say that I blame enterprise management companies for this, but it also means that as an industry we haven’t really advanced all that much since the days I used to work with NetView automation CLISTs and REXX execs at IBM in the 1980s.

  • 3 Steve Chambers   February 10, 2009 at 4:54 pm

    Cameron, I remember CLISTS and REXX execs… my my that seems a long time ago

    I’m working at VMware looking at management operations – I agree with you on the ease of describing a framework and the difficulty of executing on it… I’m working with a few others on this niggly problem and continue to meet large customers every week with similar challenges… if you are doing any research in this area I’d like to understand more!

  • 4 Cameron Haight   February 13, 2009 at 2:44 am

    Thanks for your comments Steve (and it’s good to know that there are a few of us “old timers” still around). Actually, I’m familiar with your work on the viops website, and would enjoy a discussion at your convenience as I’m still very focused on the operational impact of virtualization. Best way to arrange this would be to coordinate through your analyst relations person. If you don’t know who this is, please ping me at and I will forward you a name. Many thanks again.