AI monitoring system gets on-the-job training at ports

AI monitoring system gets on-the-job training at ports

The National Ocean Service's CO-OPS Web site posts information about conditions at ports around the country, such as these locations in New York and New Jersey.

CORMS II learns to give the right answer from missing or contradictory data

The National Oceanic and Atmospheric Administration's National Ocean Service is using artificial intelligence to monitor the quality of readings from meteorological, current and water level sensors.

By checking with the Continuous Operational Real-Time Monitoring System II, mariners can tell if a tanker should navigate a harbor, how heavy a barge can be loaded without compromising safety or if adequate clearance is available under bridges.

CORMS II, which was developed for NOS by Northrop Grumman Corp. using ARTEnterprise from MindBox Inc. of Greenbrae, Calif., checks the accuracy of sensors at harbors around the country.
The system got its start after an accident on Florida's Tampa Bay.

Blown away

'Mariners were relying on tide and current tables,' said Tom Bethem, chief of the Information Systems Division of NOS' Center for Operational Oceanographic Products and Services (CO-OPS).

'This data is sometimes inaccurate due to a wind event or an unpredicted occurrence,' he said.

The center developed the Physical Oceanographic Real-Time System to check tide, current and water level by polling sensor instruments every six minutes. Ship captains, pilots, lock and barge operators, recreational users and engineers access information about nine ports via the Web or a voice response system developed by Syntellect Inc. of Phoenix that provides digitized wind and currents readings.

Data from each port is routed to NOS' Silver Spring, Md., headquarters to check its accuracy. But as the number of systems grew, it became increasingly difficult to isolate faulty data.

When the original version of CORMS proved inadequate to handle the job, NOS decided to expand the system to view historical trends, analyze errors and detect patterns in them.

CORMS II combines rule-based and case-based reasoning. Rule-based systems apply policies, procedures and best practices to problems. As data enters an application, an engine selects the rules that apply to the current situation and uses them to take appropriate actions.

Case-based reasoning assists more complex decision-making. The engine stores a case base of solved problems and searches for similar cases to adapt their solutions to a current situation.

By combining rule-based and case-based reasoning, CORMS II can process incomplete or contradictory data to provide operators with suggested remedies.

CORMS II runs under Microsoft Windows 2000 and uses a Microsoft SQL Server database.

'The system learns from the cases and scenarios that come up, so it becomes smarter with time,' Bethem said. 'It is able to recognize potential situations, note historical patterns and provide operators with guidance.'

Decision history

When technicians at a port are dispatched to investigate a problem, they can check CORMS II to see how a decision came about and add information to help the system make better evaluations in the future.

'AI will let us take the system to two or three times as many ports,' said Bethem. 'It gives us a consistent and objective way for making decisions about the quality of the data and a rapid means of taking corrective action.'

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above