The Empire’s Bad Architecture Decisions

1 MINUTE READ
Will Claxton | November 8, 2017

I’m a father who was 5-years-old when the first Star Wars movie hit the screen.  I’m a child of Lucas and I’m really excited about the upcoming release in December:  The Last Jedi.  Also, as a father, I’ve sort of monopolized Sunday evenings for Star Wars replays (1-7) so that my two boys will understand what’s happening.  

Last weekend we watched Star Wars Episode II:  Attack of the Clones.  I don’t think my 9 year old was unique in his thoughts that the architecture the Empire had for the clones was just all wrong.  Putting all the logic for the clones up into a datacenter in the mother ship (read also “the cloud”) was a pretty bad idea.  As a technologist, I had to wonder what kind of low-latency connections those drones had to have in order to do battle in real time.  I can’t imagine API response times are that much better in a galaxy far, far away.  You get the point and probably aren’t “buying it” either.  Bad architecture.  It wasn’t included in the movie footage; however, I suspect the CTO for the Empire paid a fairly high price in his next meeting with the “top brass”.  

As a technologist, I think MachineShop’s product would have been a much better path for the Empire.  The Clones on the ground would have had “independent thought” to a much greater degree had the Empire’s CTO adopted an edge computing standard.  Don’t get me wrong...MachineShop isn’t “artificial intelligence”...that stuff still resides from our perspective up in the cloud.  However, basic logic IFTTT (if this, then this, else this) could have continued on from within the “edge devices” (see also: Clones) and in the case where the primary data center “goes dark” (or blows up) the logic at the edge could have easily re-routed the decision-making process to another location.   

Wars are won and lost...even those purely based in the business arena as well as those with live fire…based on these architecture decisions.  When looking at the edge, there are some significant architectural differences that are evolving and it’s definitely worth evaluating these if you are managing a pile of IoT devices.  



Topics: IoT, Device Management, edge intelligence, edge computing