Simple autonomous inspection vehicles are suitable for operations where the cost, danger to humans, or area of operation prohibits the use of conventional underwater technology. Autonomous vehicles are, however, in their infancy and few such vehicles are available. There are still some problems to be overcome before this technology becomes useful in commercial applications. We have built ROV90 to investigate these problems. It is a test bed for experimenting with the different parts of an autonomous underwater vehicle. ROV9D will be able to autonomously follow prominent features In the real world, man made or natural. Examples are pipelines or walls in tunnels. ROV90 is tethered, but we are planning to use experience and results from ROV9D to develop a "real" AUV called PISCIS.
Today's tethered ROVs Can carry out many of the operations for which a diver used to be necessary. The ROV's maneuverability and range are, however, limited by the tether. The necessary surface support for a UUV or RCV adds much to the cost of ROV operations and makes it dependent on good weather conditions. Thus, there is a large potential for reducing the cost of traditional ROV operation by removing the tether and if possible, the surface support. On the other hand, an untethered underwater vehicle (UUV) have to have a much more advanced control system than the ordinary ROV. The reason for this is that a hydro-acoustic link is the only practical channel for communication underwater. Operator interaction is made difficult because this type of link has low band-width, transmission delays In the order of seconds, and limited range. The UUV must have a control system that can cope with the limitations of the hydro-acoustic