Technologies for Low-Crew/No-Crew Ships

Related transport engineering sciences show spectacular progress in automation reviving a discussing of unmanned ships using Artificial Intelligence technologies particularly for combatants. The nautical tasks could be largely automated and the commercial expert systems for automatic navigation including collision and grounding avoidance are on the market. Damage control appears to be another promising area for expert systems. Natural language interfaces and gestures allow better manmachine interaction. Machine vision is seen as a frontier technology to enable further automation. Humanoid robots appear rather useless for navy ship operation, but micro-robots and simple fixed robots may be used for assorted tasks. Virtual reality is predominantly attractive as training tool. Transponders will allow automatic ship-ship and ship-shore communication. Transponders may also implanted to humans as a convenient “key” to interact with computers. 1. Technology survey in related transport engineering industries Computers take over controls in cars, trains and planes: • Intelligent highways and self-driving ‘seeing cars' Self-driving cars are not a new concept. "Indeed, a working model of an automated highway was the hit of the General Motors pavilion at the 1939 World's Fair in New York City. During the late 1950s and 1960s, researchers at General Motors went on to refine various driveless vehicles. They showed, for example, how robotic trucks could work in open-pit mines. Although these early attempts at automation were valuable research exercises, the results proved too crude to be truly workable. Yet by the late 1980s, advances in microprocessors, wireless communications and various electronic sensors prompted many people to rethink the idea of automated highways. One group, which originally called itself Mobility 2000, convened in 1988 to consider the possibilities. It subsequently formed the Intelligent Vehicle Highway Society of America (later named the Intelligent Transportation Society of America), which now has more than 1000 organizations as members. Its mission is to foster the introduction of various ‘intelligent' transportation systems, including automated highways,'' Rillings (1997). In 1997, a prototype system for an automated highway on a stretch of the California freeway showed how automation might allow existing highways to accommodate a larger number of vehicles, while ensuring a higher degree of safety. The developers target the year 2002 for a commercial introduction of the system. While the Americans base their system on sensors and signals embedded in the freeway, guiding cars along electronic tracks, German developments of ‘seeing cars’ are even more spectacular. Researchers around Prof. Ernst Dickmanns have developed self-driving cars that use video and pattern recognition to supply a computer with the necessary data to model the world outside, Fig.1, Maurer and Dickmanns (1996), www.unibw-muenchen.de/campus/LRT/LRT13. The actual control based in part on a knowledge-based systems for traffic rules and other rules for driving a car is then rather simple as cruise-control has long been established in cars. By 1994, the ‘seeing car' drove automatically in Paris on a three-lane freeway at speeds of 130km/h changing lanes and passing other cars. Also in 1994, the ‘seeing car' mastered the problem of turning into intersections. By 1997, the ‘seeing car' recognized street signs and potholes in the street and could drive 180km/h safely. Remaining problems for a commercial introduction will be overcome as computers become smaller and more powerful. It is estimated that in 20 years the necessary computer power to allow automated driving by ‘seeing cars' will have the size of a football and be installed in cars on a standard basis. Simpler systems, that automatically reduce speed if getting too close to the leading car, are already on the market.