Robots are Coming! (soon, to an enterprise near you)

Posted on October 14, 2007

Some of the most famous robots I’ve had the pleasure to observe are UAVs (unmanned aerial vehicles).  These devices usually operate under human control, but most also come with the advanced circuitry that allows them to decide how to move from point to point through the air.  This is some really complex technology that requires the UAV to make and execute decisions on how to accomplish what its human master tells it to do.   

UAVs are an interesting example, because their use is becoming more widespread.  They are not only for operation over battlefields.  They are now increasingly being used by law enforcement, firefighters, fishing fleets and academic researchers studying all sorts of atmospheric and geodetic effects.   

But UAVs, like most other robots designed by and for large enterprises, are focused on accomplishing enterprise driven missions.  Robots designed and built for individual consumers are also arising and these will likely provide tremendous value not only for individuals but for enterprises. 

For example, a great deal of robot research in Japan is focused on finding ways to help care for its aging population.   Models are for sale now that are designed to keep elderly people entertained, stimulated and interacting with their environment.  New designs are being fielded with advanced senses of smell and other sensors designed to evaluate a human’s health.   Other  robots in Japan can feed by spoon or assist the elderly in moving around.   During a trip to Japan this summer I noticed more use of robotic contraptions/devices in places like parking lots and toll booths, but also robotic traffic control.   Several Japanese companies are selling versions of robots for security, including one that looks a lot like  RoboCop.

If I’ve gotten your curiosity up, Wikipedia has a good overview of the state of robotics.  That entry is well worth a review by any enterprise IT professional, because the coming rise of robots is going to have a huge impact on our future requirements.

For example, enterprises everywhere will need IT systems that can handle increases in sensor data from robots.  And most enterprises will need new tools to help plan the optimal use and placement of robotic sensors and other robots.  And most enterprises will need help desks staffed with people who know how to troubleshoot robots (or maybe robots can staff the help desk?).  Enterprises that work with sensitive data will need to establish policies on which robots will be allowed into their spaces since every one can collect information. 

And then there is robot repair. 

To underscore the need for robot repair, I’ll close with a personal anecdote.  I saw Asimo at the 2007 Consumer Electronics Show in Las Vegas.  He is so cool!  He can walk, talk, play soccer, dance, and handle a wide range of items.  He can recognize his environment and respond to it.  It was great to meet Asimo first hand.   During his stage performance, his human handler explained many of the feats the Honda engineering team accomplished in building Asimo.  We were all impressed.  The last big stunt Asimo was to perform for us was to run in a huge circle around the stage.  This, we were told, required many onboard calculations  to be done because of the advanced calculus required.  Asimo would need to think through the size of the stage, the placement of obstacles, his own planned speed and how to turn a little at a time to make a big circle.  So, we were told, Asimo would pause for 30 seconds to make those calculations.   Well he paused.  And paused.  And paused.  Two minutes later we were told that little Asimo had totally locked up.  He needed a reboot.  The show was over.   

That made me immediately realize two things.  For one, I was able to deduce what operating system little Asimo runs.  For two, it made me realize that many of the things we have seen about robots in bad science fiction might one day come true.  There can be glitches and ghosts in the machine that may result in unpredictable behavior.  So we enterprise technologists will always need ways to re-exert our control over robots.  That must be a meta-requirement.