CASE BASED REASONING SYSTEMS FOR WINTER ROAD MAINTENANCE

Deployment of maintenance services in response to winter conditions requires accurate decision making, often based on incomplete or probabilistic information about developing storm systems, existing avalanche potential, etc. To aid decision-makers, computerized decision support systems can be developed. Most of these computerized aids are rule-based systems which require specific data parameters. If that information is not provided or is inaccurate, then the computer system fails to operate or gives inaccurate information. An alternative to such rule-based systems is a case-based system which stores prior maintenance instances (cases) and draws upon these cases to determine the similarity of the current conditions to those of the stored cases. The advantages of such a decision aid and the methods for developing one are discussed here.