Expert systems, also known as "knowledge-based systems," are a class of computer
program that can advise, analyze, categorize, communicate, consult, design,
diagnose, explain, explore, forecast, form concepts, identify, interpret, justify,
learn, manage, monitor, plan, present, retrieve, schedule, test, and tutor.
These types of systems attempt to remove some of the complexity and skill
necessary for certain jobs. Typically, an expert system is capable of
explaining its reasoning for the decisions it makes and can "learn" as it
gains experience. Of course this learning is not like a human learns. Expert
systems will instead employ algorithms and store data using the approaches
described in any Artificial Intelligence book.
Benefits of Expert Systems
- Reduced error due to automation of tedious, repetitive or critical tasks
- Reduced manpower and time required for system testing and data analysis
- Increased visibility into the state of the managed system
Capabilities of Expert Systems
- Develop functional system requirements
- Coordinate software development
- knowledge acquisition, process analysis, data analysis, system verification
Design, develop, and implement an intelligent system
In Scandanavia, the airlines had attempted to design systems that would automate much of
the maintenance process for airplanes. Essentially, the expert system would be able to
diagnose potential problems with parts of the planes and provide complete and detailed
information about the course of action a mechanic should take. Unfortunately, as noted
in the film Computers and Context, the mechanics relied on the computer too much.
Because the suggestions came from a computer, the workers felt less confident about
their own personal evaluations of the situations and tended to place complete trust
in the system.
With Participatory Design, the airlines were able to implement a system that
augmented the skills necessary for workers to maintain the airplanes, rather than
attempting to completely replace the skilled workers. This new system was much better
described as a "System for Experts." This is the approach Participatory Design attempts
to take in creating systems. In the case of the Scandanavian airlines, it is a better fit than attempting to create systems that replace the skilled workers.
Often system designers fail to take into account the actual users of the programs
that are being implemented. There is an attempt by programmers to incorporate too many
features or too much functionality in an attempt to make the user more productive.
Unfortunately this methodology may make programs more complex and actually decrease
productivity as the user attempts to sort out the actual functionality he or she
desires. These types of program also tend to not take into consideration the way
the users actually work. Participatory Design allows the users to give input into
the design process and make it more likely that the finished product will have the
functionality required and work as the users envisions.
Ultimately, Expert System Designs are much better suited for situations where there
is already a relatively unskilled labor force attempting to do a job. If a system is
being built for skilled labor, Participatory Design is much better at helping to extend
the capabilities and productivity of a worker. Expert Systems are perfect for sensor
validation jobs where many different sensors monitoring conditions or for repetitive
jobs that are much better suited for automation. In these types of situations, a computer
is better at performing the task. For other tasks, new systems need to enhance the
ability people already possess.