The problem of seeing things underwater is one that has vexed divers and others who work in the water since time immemorial. Ordinary vision works OK, in the daytime or with artificial light, in shallow water, in good conditions – but fails spectacularly in poor light or when water conditions are murky or worse. The modern electronic marvel that let us “see” for thousands or even millions of miles through air and space – radar – is completely useless underwater; water is effectively a brick wall to the short electromagnetic wavelengths that radar employs. Yet it is extremely important that underwater craft be able to sense their surroundings; if your craft is unable to see what is around it, you will pay the price for your lack of vision. Around the beginning of the 20th century, researchers realized that they could use sound waves to “see” underwater.
Unlike radar, sound waves propagate just fine underwater – in fact, they propagate underwater better than they propagate through the air. The first sonar-like devices were used to listen for icebergs; sonar technology advanced enormously during the First World War as a tool to listen for enemy submarines. Sonar still has extensive military uses, but today civilian sonar is the primary area of development.
Although in this post we will talk about “seeing” it should be noted that sonar does not actually produce a direct visual image. Many sonar control suites will translate the sonar data into a visual picture, and high-end sonars actually can produced simulated images that look very much like you are “seeing” what is out there – but the actual data is sound pulses being reflected back. Computers do much of the interpreting of that data stream, making it more accessible to people without advanced training in sonar interpretation.
Most remote-operated vehicles have the ability to carry sonar sets as optional equipment.ROVs use sonar for a variety of purposes, from mapping the bottom of bodies of water to searching for items in the water.
There are three basic types of sonar available for deployment on an ROV:
Scanning sonar operates in a way that will be familiar to anyone who has ever seen a radar dish turning at an airport. The sonar emitter physically rotates, sending out a pulse of sound as it does so, and simultaneously listens for the echoes returning from any objects in the water. The rotational speed of a scanning sonar balances out the time between pulses and the time it takes for a returning pulse to come back to the sonar.
Scanning sonars do not provide a high degree of resolution. They are able to detect large objects (for example, pilings or ships) but lack the discrimination necessary to spot smaller objects, particularly when those objects are on the bottom or next to another object. Because there is only one beam, scanning sonars have no ability to see behind objects; if one object is behind another object in a line to the scanning sonar, the scanning sonar can see only the closer of the two. The strength of the scanning sonar is that it does cover an area of 360 degrees, although only from 10 to 20 degrees above and below the plane of the sonar.
Multibeam sonars (often referred to as “hydrophones”) are significantly more capable. A multibeam array has multiple sonar emitters, which are all digitally rather than manually moved, allowing for a much higher rate of scanning. A combination of hardware and software allows the multibeam sonar to sweep a wider area of the water – about 120 degrees in all directions. Multibeam sonars do not cover a 360 degree arc, however. They make up for this by having a higher degree of resolution than scanning sonars – at a hundred meters, a multibeam sonar can spot a diver in the water. At 10 meters it can distinguish arms and legs. Multibeam sonars are extremely good at looking in one direction at a time.
Side-Scanning sonars have two arrays, each of which has a sonar beam that traverses a 90-degree arc, one horizontally and one vertically. This means that they can see in almost every direction, except for straight up and down. Side-scanning sonars are used to collect data as an ROV moves through the water; unlike scanning and multibeam sonars, they do not provide information in real-time. Instead, they provide a “historical” view of what was there when the ROV was moving past a certain point. The advantage of side-scanning sonars is that they “see” almost everything, providing great resolution and covering an enormous area. Side-scanning sonars are suitable for wide-area searches for stationary objects (such as bodies or objects on the bottom of the body of water), surveying, and environmental monitoring.
The right type of sonar for your ROV will depend on the missions for that ROV. Each sonar type has advantages and disadvantages for various kinds of work. Aquabotix ROVs can be configured with any of these sonar types. As with other accessories, choosing the right tools for the job will go a long way towards ensuring that you get the best possible results.
One of the important elements in ROV mission performance is the quality of the visual information that the ROV passes back to the operator. Sending an ROV into the water lets a human being put eyes on the target, and those eyes need every well-lit pixel they can get.
What lets an ROV operator get the highest-quality underwater visual information back from his or her vehicle?
A significant component of video quality is the video format being used by the camera. For most of the 20th century, video cameras used either the NTSC or PAL standard. NTSC displays had (and still have) 480 lines of usable vertical resolution, while PAL used 576 lines. Both the NTSC and PAL formats had 4:3 pixel ratios, meaning that an NTSC picture is 640x480 (not coincidentally, the same screen size as the old VGA computer monitor standard) while a PAL picture is 758x576. These resolutions were adequate for the black and white and color TVs found in most homes up until around 2010.
Both PAL and NTSC are interlaced formats, meaning that every other scan line is refreshed at each frame (i.e., half the picture is refreshed in one frame, the other half on the next frame). This produces a smoother picture at reasonable frame rates, with the unfortunate downside of producing jitteriness during motion shots. Neither NTSC nor PAL would make good formats for underwater video for this reason, even if their low resolution was not an obstacle: an NTSC frame has only 307,200 pixels and a PAL frame has only 436,608.
Modern video cameras use one of the HD (high-definition) standards. The lowest form of HD is 720p, which provides 720 lines of vertical resolution with a progressive (non-interlaced) scan. This provides 921,600 pixels, which is a step up from NTSC or PAL but still leaves something to be desired.
The midrange HD standard is 1080i, which has 1080 lines of vertical resolution but unfortunately reverts to an interlaced display. This provides 2.07 megapixels per frame, but only 1.03 megapixels refresh with each frame, leaving the video subject to jerks and jittering. The highest HD standard is 1080p, which provides a full 1920x1080 2.07 megapixel frame without interlacing. 1080p is the best commercially-available picture format; 1080p video at 24 frames per second is the equivalent of watching a Blu-Ray DVD on an HDTV. 1080p at 24 frames per second is the gold standard for underwater video quality.
However, size is not everything, and there are other important elements to consider when maximizing underwater video quality. One critical area is lighting. In many ROV applications, such as water tank inspection, there is no natural light at all, and even in open water the sunlight diminishes drastically with depth. At 10 meters down in crystal clear water, there is only about 25% as much light as at the surface and at 100 meters down there is only 0.5% as much light. Rough water and particulates in the water can make things even dimmer. Even if you have adequate natural light for your purposes, remember that water causes a distortion in the color spectrum; red light is absorbed most rapidly causing things underwater to tend strongly towards appearing blue or green. If your application needs accurate color, or if you need more light than is naturally available at your depth, then artificial lighting from the ROV itself will be necessary to get good video quality underwater.
The optical characteristics of the lens and the mechanical elements of the camera itself are also important in underwater video work. A wide-angle lens allows a very close focus on objects in the field of view, which means that the operator can get very close to the subject and still have a clear picture. A camera that can tilt and pan while the ROV hovers in one place can get even tricky shots into close focus. Of course, the maneuverability and ease of control of the ROV itself can be determining factors here – you have to be able to get the camera to the target, so an ROV that is easy to control is one from which it will be easy to get good video. The combination of a high-quality video format, a high-yield lighting array, a well-designed optical system, and a maneuverable, flexible ROV will provide the best possible underwater video quality for any application.