Start Date
4-1987 8:00 AM
Description
Computers are being used today to build the expert systems of tomorrow. Expert systems are computer programs that are smart about a domain in the way that people are smart. Expert systems technology is being applied to weather forecasting to support Shuttle operations for launch and for ground processing at Kennedy Space Center (KSC), Florida. The Space Transportation System Meterological ExperT (STSMET) is a long term project, now-in its third year, to capture general Shuttle operational weather forecasting expertise specific to our locale, to apply it to Shuttle operational weather forecasting tasks at the Cape Canaveral Forecast Facility (CCFF) at the Cape Canaveral Air Force Station (CCAFS), and to ultimately provide an on-line, real-time operational aid to the duty forecasters in performing their tasks.
The first domain addressed by the project has been summer thunderstorms. The effort to represent this knowledge and a control structure to reason about it has resulted in an approach that we call scenario-based reasoning. Other meteorological domains on our agenda are frontal weather phenomena, visibility including fog, and wind shear. We believe that scenario-based reasoning is also applicable to these other meteorological domains. The specific operational tasks to which to apply the general knowledge about summer thunderstorms are being identified during this phase of the contract.
The project is being developed using state-of-the-art hardware and software: a Symbolics Lisp Machine, Zetalisp and Automated Reasoning Tool (ART), an expert system shell.
Scenario-based reasoning appears to have applications outside of weather forecasting. The abilities of a scenario-based system to reason qualitatively, to reason over time, and to reason across scale are all applicable to planning in autonomous systems. With further research, we expect to add analogical reasoning to the abilities of scenario-based reasoning.
Space Transportation System Meteorological Expert
Computers are being used today to build the expert systems of tomorrow. Expert systems are computer programs that are smart about a domain in the way that people are smart. Expert systems technology is being applied to weather forecasting to support Shuttle operations for launch and for ground processing at Kennedy Space Center (KSC), Florida. The Space Transportation System Meterological ExperT (STSMET) is a long term project, now-in its third year, to capture general Shuttle operational weather forecasting expertise specific to our locale, to apply it to Shuttle operational weather forecasting tasks at the Cape Canaveral Forecast Facility (CCFF) at the Cape Canaveral Air Force Station (CCAFS), and to ultimately provide an on-line, real-time operational aid to the duty forecasters in performing their tasks.
The first domain addressed by the project has been summer thunderstorms. The effort to represent this knowledge and a control structure to reason about it has resulted in an approach that we call scenario-based reasoning. Other meteorological domains on our agenda are frontal weather phenomena, visibility including fog, and wind shear. We believe that scenario-based reasoning is also applicable to these other meteorological domains. The specific operational tasks to which to apply the general knowledge about summer thunderstorms are being identified during this phase of the contract.
The project is being developed using state-of-the-art hardware and software: a Symbolics Lisp Machine, Zetalisp and Automated Reasoning Tool (ART), an expert system shell.
Scenario-based reasoning appears to have applications outside of weather forecasting. The abilities of a scenario-based system to reason qualitatively, to reason over time, and to reason across scale are all applicable to planning in autonomous systems. With further research, we expect to add analogical reasoning to the abilities of scenario-based reasoning.
Comments
No other information or file available for this session.