CHERRY HILL, N.J. — Teams of sailors at the maritime security operations center in Italy monitor ship traffic plying the Mediterranean Sea searching for watercraft that are acting suspiciously. But discerning the potentially hostile ship amongst the tens of thousands of benign vessel tracks can be a difficult, time-consuming process.
A Web-based software program under development aims to make the task easier by giving watch teams the tools to automatically monitor, track and analyze pertinent events occurring in the world’s oceans and waterways. “It’s useful for turning that mountain of data into actionable information quickly,” said Rich Dickinson, program manager for the maritime agent analysis toolset at Lockheed Martin Corp.’s Advanced Technology Laboratories.
The software allows the teams to set up and activate “agents” that can help sailors hunt down specific information culled from a variety of military and commercial databases. The commercial databases include the automatic identification system live, a Surrey, United Kingdom-based global network that disseminates the real-time transponder signals that are transmitted by ships.
The maritime agent analysis toolset has a number of preprogrammed templates that operators can configure to help them zero in on ships that are behaving suspiciously. Templates such as abnormal vessel speeds, vessel or geographic proximity, characteristic change, and direction of movement will launch an “agent” that will search for violators within those specified parameters. If it gets a hit, the software lists the ship. Operators can then click on the list to see more details about the vessel.
“The agents are your eyes and ears,” Dickinson said. “They’re watching for the various conditions.”
Sailors can pull up the ships on a Google Earth display to see where they are located and where they have been and where they are heading. The software was derived from an earlier Defense Advanced Research Projects Agency project.
In a demonstration, Dickinson logs onto MAAT and pulls up the Google Earth representation. He zooms in until it shows the nearby Delaware River and a number of green, yellow, and red icons, each representing different vessels that are color-coded by how recently they have reported to the AIS system. Green icons have reported within the last hour, yellow icons have transmitted within the past two to 24 hours, and red icons are those that have not signaled for more than 24 hours.
Dickinson clicks on a vessel that he had begun tracking a few hours earlier. It is sitting at the mouth of the Delaware waiting for the pilot to come and take it up river. A quick click makes the system show the previous 250 positions for the ship. It lays down the track and an operator can see where the ship has sailed from.
“You couldn’t do this just by looking at the [common operating picture],” said Dickinson. An operator could sit and watch the vessel tracks coming into his area, but he wouldn’t know the speed, or where the vessel had been two hours earlier, or how long it had stayed put in one location, he explained.
He clicks out of the Google Earth visualization and over to a menu with active agents. He pulls up one that was set up to monitor ships moving slowly near the airport. In the course of an hour, the agent has tallied 26 hits. “That agent is poring through the tomes of data so you don’t have to,” he said. “It is vigilantly watching the data feed for that condition that you specify as part of your configuration process.”
Dickinson selects the number 26 and a list of the violating vessels appears. There are four ships, of which the Reed McAllister has the greatest number of occurrences with 12 hits.
Agents can also search through archival data. If an operator wanted to know where Reed McAllister had been four days prior to arriving in the Delaware, the agent would bring up the ship track and display its tail through Google Earth.
The software can find data quickly and give operators more time to do the analysis, he said. It also provides tireless, round-the-clock vigilance.
“These things will look for those conditions and won’t miss it, whereas a human may not be able to recognize the certain condition they were looking for,” or he misses it altogether, Dickinson pointed out.
Though the lab designed the program as a shore-based product, researchers are discovering that it could have utility deployed on board a ship as well. Last summer, the team participated in an experiment with the Joint Interagency Task Force-South, which is leading U.S. counternarcotics efforts. To test out sensors being developed to detect small watercraft, the task force staged mock scenarios with go-fast boats simulating drug smuggling runs. The MAAT software was placed aboard a ship whose radar data was fused with its AIS receiver. The combined information was sent to a collection center in Minnesota to help fill in the gaps that weren’t covered by the sensors. By placing MAAT’s “bubble” agent around the vessel, the ship’s crew could track the go-fasts when they came within a quarter of a mile of their location.
Dickinson said that the team is looking to participate in the military’s Trident Warrior exercise this summer.
The software is applicable to other domains, and lab officials said that it is being packaged as a product for tracking spatial and terrestrial objects.