Enable breadcrumbs token at /includes/pageheader.html.twig

Robo Boats On the Horizon

Autonomous unmanned vessels soon may be patrolling harbors and conducting coastal reconnaissance missions. The prototype for these future robo boats can operate cooperatively with other robotic surface craft and navigate to a destination without human guidance. Designed for use in shallow waters, the boat can connect to and be controlled by U.S. Navy ships via tactical communications networks.


The AMN 1 unmanned surface vessel (USV) can operate autonomously and in cooperation with other robotic vessels. During the recent Trident Warrior 09 exercise, the USV patrolled the waters around the USS Nassau. The robot vessel detected and interdicted boats approaching the warship entirely in autonomous mode without any human control.

Smart unmanned craft may replace warfighters for both dangerous and routine operations.

Autonomous unmanned vessels soon may be patrolling harbors and conducting coastal reconnaissance missions. The prototype for these future robo boats can operate cooperatively with other robotic surface craft and navigate to a destination without human guidance. Designed for use in shallow waters, the boat can connect to and be controlled by U.S. Navy ships via tactical communications networks.

Known as the AMN 1, for autonomous maritime navigation, the unmanned surface vessel (USV) recently participated in the Navy’s Trident Warrior 09 exercise. The USV operated some 1,000 yards away from the amphibious assault ship USS Nassau (LHA-4) and performed force protection exercises by interdicting a target boat approaching the warship. During the exercise, the USV operated in autonomous mode with no human remote operation. For safety reasons, a boat operator and a technician were present on the USV, but they did not control it.

The AMN 1 evolved from an earlier program, the Combat Marine Vehicle effort, explains Eric Hansen, a program engineer with the Naval Network Warfare Command’s Combatant Craft Department, Norfolk, Virginia. The AMN 1 effort currently is funded under the Griffin program. Launched in 2007 with $3 million in funding, the Griffin program is part of a rapid technology effort seeking to develop USVs for cooperative group operations. The program has achieved more than this goal, he says.

However, Hansen notes that the program currently operates in a vacuum because the Navy has not released any operational requirements for USVs. He explains that some of the concepts about autonomous robotic operations originally were developed a decade ago, when U.S. Defense Department planners began considering the possibility of robots working cooperatively in swarms. Autonomy was seen as an important feature to allow robots to cooperate in the field. But military and civilian robot development has always proved to be more difficult than original projections. He notes that it has taken a while for the technology to catch up with the promise and potential of military robotics.

The AMN 1 is a purpose-built USV, not a modified commercial vessel. Program engineers sought to make an affordable, easy to maintain, general platform that could carry a variety of payloads, Hansen explains. The USV is manufactured by Oregon Iron Works Incorporated, Clackamas, Oregon, which designs and produces military patrol craft. He explains that there is nothing exotic about the robotic vessel other than its specialized design. The AMN 1 also features a flat-bottomed hull and is designed for ruggedness and shallow-water operations.

Hansen says that the USV can perform a variety of missions such as harbor security, river operations and intelligence, surveillance and reconnaissance missions. The AMN 1 also can be used for homeland security and law enforcement operations including search and rescue, interdiction of vessels and maritime patrol missions. Hansen adds that the USV would be useful for certain types of reconnaissance and interdiction missions that are dangerous for human personnel. The craft autonomously can approach a suspect vessel; scan it with sensors; and compare the craft against an onboard database of suspected pirate or narcotics trafficking vessels. If a vessel is positively identified as a threat, the USV will alert nearby warships or harbor authorities.

To navigate, the AMN 1 uses lidar, a K/A band radar and a 360-degree wide-field stereo electro-optical system. The craft also uses Global Positioning System navigation, and it has been equipped with side-scanning sonar for specific missions. Hansen says additional plans are to install stereo infrared sensors to the vessel.

All of these onboard sensors can be fused to create a common operational picture. The USV’s operators also can select combinations of sensors to focus that provide the most information for specific missions. Hansen adds that sets of sensor packages optimized for individual operations can be set aside, installed or activated on a per-mission basis.

The AMN 1’s autonomous navigation technology is based on the control architecture for robotic agent command and sensing (CARACAS) capability designed to guide NASA space probes and planetary rovers. CARACAS is a resource-based autonomy system that focuses on preserving the safety of the vehicle and managing resources such as fuel and battery power. Hansen explains that this capability is effective for ships at sea where, because of the size of the ocean, collision avoidance is secondary to managing internal systems such as fuel, propulsion, communications and flotation.

Because CARACAS originally was designed for NASA, Hansen describes it as “taxpayer-owned.” He envisions a business model where the government owns the system end-to-end “like Apple owns the iPod,” he says. The navigation system features a behavior tool kit. Third-party developers can write applications for new payloads or operational behaviors that are then added to the USV’s navigation system in a manner similar to developers loading applications into an iPod.


The navigation system for the AMN 1 USV is based on software designed to provide NASA space probes with autonomy. The robot boat can be controlled remotely, or it can be given the coordinates for a location and can autonomously select the best course to follow.

Hansen developed this business model because of certification and validation concerns about third-party software. Because of the risks associated with software changes, the core system that governs the vessel’s vital operating functions is owned and maintained by the Navy.

The USV can be teleoperated or it can follow waypoints. The vessel can be given a specific goal, and it autonomously can determine the best route to the objective. Because changes can happen mid-operation, the navigation system also allows the robot boat to re-plan its mission in transit to the target.

Developing systems to navigate autonomously in a maritime environment has proved to be more complex than expected, Hansen says. Perception is the key issue. Autonomous ground vehicles deploy a variety of sensors because they must deal with many objects, such as bushes and rocks, which can impede their travel. He notes that during the Defense Advanced Research Projects Agency’s first Grand Challenge robotics contest, some teams measured and recorded every object along the vehicle’s path. But such an approach does not factor in random issues. In that first Grand Challenge, the design teams measured everything except tumbleweeds, which are mobile and confused the robots.

Maritime environments differ from land because they can be very dynamic. Hansen says that an important aspect about the sea is waves, which create reflections. These reflections, caused by sunlight on foam and spray, affect both optical and radar systems. The boat’s movement also affects how the sensors operate and relate to their environment. Another issue was navigating around some objects. He explains that the team’s engineers had difficulty programming the AMN 1 to travel under bridges. However, the nautical environment also has some advantages in that events tend to happen more slowly and can be detected at greater distances, he says.

Nautical distances also affect the sensors. Hansen notes that the USV’s lidar system has an effective range of less than a mile. Spray and mist also reduce the effective distance of optical systems because the boat is close to the surface and does not enjoy the height advantage of a ship’s superstructure.

Hansen’s office is actively developing a mine warfare payload for the AMN 1. Work is underway on launch and retrieval systems to manage the USV from a ship, as well as control systems for additional sensor and possible weapons payloads.

The USV can be operated from a variety of control stations. Hansen notes that the vessel was designed to be adapted to accept a range of command standards such as the NATO standards agreements. He adds that the Trident Warrior exercise demonstrated that data can be passed between the AMN 1 and Navy warships and that any ship connected to the network can command the robot.

The vessel has not been tested with weapons. Hansen explains that developing a weapon mount will be the jurisdiction of any group developing a third-party payload for the vessel. He adds that his primary goal is to develop the craft so that it can operate seamlessly across a variety of missions. The choice of armed operations should be made to save sailors’ lives by sending the USV on hazardous operations missions. Hansen believes that forces deploying robots will prevail in future conflicts and maintains that the United States must have this capability.

The next goal for the USV effort is to move to a program of record. The effort currently has two cooperative vessels that will participate in a series of exercises in the Washington, D.C., area late this year or early in 2010. Hansen explains that another reason for participating in events such as Trident Warrior is to increase recognition of the system within the Navy.

U.S. Naval Network Warfare Command: www.netwarcom.navy.mil