The APM Control System

The APMs operation are totally under the control of a central computer.  The movement of every carrier and location of each individual passenger within the entire network is constantly monitored.  The decision on the next destination of each carrier that has reached its final destination hub is made based on decisions made by the central computer.  

Computer technology continues to increase exponentially.  All predictions support Ray Kurzweil’s concept of singularity which predicts that artificial intelligence will surpass human intelligence in the foreseeable future.  We have seen a dramatic example of this level of intelligence in the development of the IBM Watson that defeated some of the most knowledgeable people in the world in the in the game “Jeopardy”.  

Jeopardy with Watson in CenterThe rapid growth of software capability is often related to how software analyzes existing data to predict trends and to make decisions based on these trends.  In the IBM Watson project, the computer software had limited human programming.  Every question ever asked on a Jeopardy program since the show initially aired in 1984 was fed into Watson.  In addition, every database that had information that could be useful to answer these question was also loaded into the computer.  The software used self-modifying code generation to constantly adjust its predicted answer to match the correct answer.  In situations where the software could not provide the correct answer, the programmers would write “rules” that would provide directions to the correct solution. This is the basis of how the data mining software at Facebook, NSA, National Weather Service, Google etc. operate.

The APM is the perfect application for this new generation of software. The APM will have sensors that continually monitor carrier movement in the hubs, read passenger RFID cards throughout their travels within the network, monitor equipment performance, etc. This information is fed to the network computer via fiber optics lines that interconnects all of the hubs. Once the software has determined the cabins destination, its movement and passenger load can predict its movement along its path with accuracy.  As this carrier travels, its precise location is identified by the physical sensors.  The software determines if the carrier and passenger movement meets an acceptable tolerance.  This tolerance is determined by historical data.  The software reacts only if the tolerance is exceeded.  The reaction is based on predictable consequences and “rules”.   

The goal of the APM software is to optimize the capacity of the system based on criteria established by the programmers.  Over time the amount of data being analyzed grows and the decisions made by the system improve.  Fortunately, the operation of the APM network can be simulated before the system goes operational by the use of simulation software that can predict years of theoretical operation based on a variety of planned and predictable events.

The APM is the only mode of surface transportation that can use this new generation of software. In the APM system there are few outside variable that will influence the operation of the system.  It is not affected by weather or surface congestion. The APM is an “aerial people mover” that performs like a machine with little human operator interaction to move people in a predictable manner.