Expert systems are of most value in organizations that possess a high high-level of know-how experience and expertise. Most corporations fail to capture this corporate IQ, thereby making it available to other employees.

Expert systems are designed to receive, hold and use this corporate intelligence and provide this knowledge to other members of the organization for problem-solving purposes.

Types Of Problems Solved

Medical, legal and other professional problems are ideal candidates for expert systems. Typical problems may include:

accounting - categorizing entries against tax codes
medicine - diagnosing skin lesions
process control - adjusting product flow based upon entry variables and output requirements
financial service - trading systems logic
production - supply and demand planning systems, modular factory built homes
human resources - resource evaluation and fit

Expert systems are used by NASA to plan the maintenance of a space shuttle in preparation for its next flight. They are also used across various industries. For ease of classification, expert systems can be defined in seven main classes of use.

Diagnosis and Troubleshooting Devices and Systems

Expert systems that deduce faults and suggest corrective actions for a malfunctioning device or process. This may be applied in medical diagnosis or computer application diagnosis. The diagnostic problem must be capable of being stated in the abstract: given the evidence presenting itself, what is the underlying problem/reason/cause?

Planning and Scheduling

Expert systems in this class use a process whereby they:

analyze a set of one or more potentially complex and interacting goals
determine a set of actions to achieve those goals
and/or provide a detailed order to peform those actions, taking into account personnel, materiel, and other constraints.

This type of expert system is commonly used today in airline scheduling [flights, personnel, and gates], manufacturing job-shop scheduling; and manufacturing process planning.

Design and Manufacturing

These systems assist in the design of physical devices and processes. Their complexity varies widely, ranging from high-level conceptual design of abstract entities down to factory floor configuration of manufacturing processes.

Configuration of Manufactured Objects from Subassemblies

Configuration is where a solution to a problem is synthesized from a given set of elements related by a set of constraints. Configuration based expert systems have to date proven one of the most important; being used in the manufacture of semi-custom minicomputers, modular home building, manufacturing, and other problems involving complex engineering design and manufacturing.

Process Monitoring and Control

These expert systems analyze real-time data from physical devices with the goal of noticing anomalies, predicting trends, and controlling for both optimality and failure correction.

This class of system is most commonly found in real-time systems that actively monitor processes in the steel making and oil refining industries.

Financial Decision Making

With the complexity of factors determining the value of finanancial instruments, it is not surprising that the financial services industry has been an enthusiastic user of expert system techniques. Financial markets commonly use expert systems in foreign exchange trading.

Advisory programs are also used to help bankers decide whether to make loans to businesses and individuals and by insurance companies use expert systems to assess risk profiles to determine a price for insurance.

Knowledge Publishing

Knowledge publishing is still relatively new, but the potential for the use of expert systems in this area is evident. Expert systems best deliver knowledge that is relevant to the user's problem, in the context of the user's problem.

Two current expert systems in this class include:

an advisor which counsels a user on appropriate grammatical usage in a text
a tax advisor that supports a tax preparation program to advise the user on tax strategy, tactics, and individual tax policy.

Artificial Intelligence

To date, the use of expert systems in AI research has been limited. Most are extremely narrow in their domain of knowledge, thus making validation problematic.

Early expert systems were prone to making errors that humans would easily identify. Most were little more than elaborate decision making systems already in use.

Software Troubleshooting Systems

Try the so called expert system used for troubleshooting in Microsoft Windows. You will find this in the "help" section in the menu taskbar. Microsoft has designed their expert system to provide solutions, advice, and suggestions to common errors encountered throughout using the operating systems. In practical use, it offers little value.

Computer Video Game AI

Artificial intelligence did not really start to evolve to levels of any significance until computer games. For example, a sports video game would generally gather highly detailed simulations of game strategies from several leading sports team managers.

When a human played the game against the computer, the computer queried the 'manager' strategic logic to help make a decision on what strategy to follow. These choices inherit a certain amount of randomness that is part of the natural system, decided based on probabilities supplied by 'the experts'.

NEXT: How Expert Systems Work

Amazon.com Widgets