CARACaS was originally developed by NASA as part of an effort to develop architectures for controlling either a single autonomous robotic vehicle or multiple cooperating but otherwise autonomous robotic vehicles. In a 2008 brief on CARACaS, JPL engineers noted that the software was potentially applicable to diverse robotic systems that could include aircraft, spacecraft, ground vehicles, surface water vessels, and/or underwater vessels.

CARACaS is intended to satisfy the need for two major capabilities essential for proper functioning of an autonomous robotic system: a capability for deterministic reaction to unanticipated occurrences and a capability for re-planning in the face of changing goals, conditions, or resources.

CARACaS’ Perception Engine consists of a 360 electro-optical system with an automated target recognition system called the Contact Detection and Analysis System (CDAS); a stereo electro-optical infrared (EOIR) system; a radar and automatic identification system (AIS); and a sensor data fusion engine developed by Daniel Wagner Associates.

The Navy has extensively tested CARACaS on USVs and Unmanned Underwater Vehicles (UUVs) over the last eight years. CARACaS is now being adapted through ONR support for autonomous swarming.

ONR also requested that JPL design, integrate and test a Day/Night Hazard Avoidance System (DNHAS) for autonomous navigation of a USV under daylight and nighttime operating conditions, with CARACaS as the autonomous control system.

The new sensing system for this project is based on a previously designed wide-baseline, high-resolution stereo system that has been used on US Navy USVs for the last six years. The new system uses higher resolution cameras for a greater look-ahead distance, coupled with stereo infrared and color cameras for the operation during nighttime.

As autonomy and unmanned systems grow in importance for naval operations, officials successfully demonstrated the new autonomous swarming technology—enabled by sensors and software—over two weeks in August on the James River in Virginia.

This networking unmanned platforms demonstration was a cost-effective way to integrate many small, cheap, and autonomous capabilities that can significantly improve our warfighting advantage.

—Adm. Jonathan Greenert, chief of naval operations

In the demonstrations, as many as 13 Navy boats operated using either autonomous or remote control. First they escorted a high-value Navy ship, and then, when a simulated enemy vessel was detected, the boats sped into action, swarming around the threat.

The James River demo was enabled though discoveries in artificial intelligence, machine perception and distributed data fusion. In the future, the capability could scale to include even greater numbers of USVs and even to other platforms, including unmanned aerial vehicles (UAVs).

The new technology will allow the USVs to detect, deter or destroy attacking adversaries. Any weapons fire from the USVs would need to be initiated by a Sailor supervising the mission.

This multiplies combat power by allowing CARACaS-enabled boats to do some of the dangerous work. It will remove our Sailors and Marines from many dangerous situations—for instance when they need to approach hostile or suspicious vessels. If an adversary were to fire on the USVs, no humans would be at risk.

—Dr. Robert Brizzolara, program manager at ONR

Naval leadership has emphasized a blended force of manned and unmanned systems in recent years. Not only can USVs take on dangerous missions, thus protecting the warfighter, but even multiple USVs are a fraction of the cost of a single large manned ship.

The swarm demo announcement comes near the anniversary of the terrorist attack on USS Cole (DDG-67) off the coast of Yemen. In that October 2000 attack, a small boat laden with explosives was able to get near a guided-missile destroyer and detonate, killing 17 Sailors and injuring 39 others.

While the attack on Cole was not the only motivation for developing autonomous swarm capability, it certainly is front and center in our minds, and hearts. If Cole had been supported by autonomous USVs, they could have stopped that attack long before it got close to our brave men and women on board.

—Chief of Naval Research Rear Adm. Matthew Klunder

Resources

• Scott Savitz et al. (2013) “U.S. Navy Employment Options for UNMANNED SURFACE VEHICLES (USVs)” (RAND report)

• Naval Research Advisory Committee Report: “How Autonomy Can Transform Naval Operations” (October 2012)

• Terry Huntsberger, Gail Woodward (2011) “Intelligent Autonomy for Unmanned Surface and Underwater Vehicles” OCEANS 2011, vol. 1, no. 10, pp. 19-22

• Wolf, M. T., Assad, C., Kuwata, Y., Howard, A., Aghazarian, H., Zhu, D., Lu, T., Trebi-Ollennu, A. and Huntsberger, T. (2010) “360-degree visual detection and target tracking on an autonomous surface vehicle” J. Field Robotics, 27: 819–833. doi: 10.1002/rob.20371