Designing an underwater drone is surprisingly hard
As there has been a recent growth in autonomous flying drones, driven by advances in technology and economies of scale, thoughts turn to underwater drones becoming as cheap and ubiquitous as their airborne counterparts.
Guest blog by Mirko Bernacchi, Mouser Electronics.
At first glance, the answer would seem to be ‘Yes’. Surely the basic concepts behind an unmanned aerial vehicle (UAV) could be easily adapted for use underwater? Underwater drones seem to present similar challenges to the airborne variety. Both need to navigate and move in three dimensions, and send data back to their human controller. An underwater craft has the additional advantage of buoyancy – making it able to hover in place with the motors off. In this respect it’s closer to an unmanned airship than a typical quadrotor drone.
However, the unfortunate truth is that making an underwater drone is far from plain sailing. Underwater drones face substantial challenges, because many navigation and communication technologies that we take for granted are ineffective in water.
GPS, mobile (3G, 4G), WiFi, and radar are among the technologies that are essential for airborne drones to operate, yet are effectively useless underwater. Optical navigation (such as Lidar) and communication are also ineffective, limited to a few dozen meters at best, depending on water clarity. In fact, the range of the electromagnetic spectrum that is useful for underwater drone developers is impoverished, compared to the wealth of bandwidth available for vehicles operating in the open air.
Radio waves travel poorly through water, and even worse through an electrical conductor such as salt water. Lower frequencies penetrate better than high frequencies. But at 27MHz, the lowest frequency available for public use in most countries (perhaps best known for its usage in Citizens Band and amateur radio), underwater range is less than 3m.
Very low frequency (VLF) and extremely low frequency (ELF) signals do have a useful range, from hundreds up to tens of thousands of kilometers. But these frequencies can only be used for one-way communication from the controller to the drone, as they require power-hungry transmitters with antennae on the scale of kilometers. Moreover, long wavelengths naturally offer extremely low bandwidth – at best, 1-2kb/ps over distances of tens of meters for VLF.
Navigating and Communicating by Sound
The most suitable wireless communication and navigation technologies for autonomous drones are acoustic. Unlike radio and light, sound can penetrate water over distances of many kilometers – hence its use for underwater navigation and communication by animals such as whales and dolphins. Even with acoustic technology, signal diffusion, low bandwidth and high power consumption still present challenges.
Currently, the best available acoustic underwater modems offer bandwidth below 20kb/ps over ranges of tens to hundreds of meters. This falls far short of the bandwidth required for real-time video. A human operator on the surface can control a drone via acoustic signals, but will be flying blind, unable to see what the vehicle sees.
As well as providing a communication channel, sound waves offer a means of navigation. Sonar can provide a real time view of the seabed and submerged objects, at ranges up to thousands of meters for low-resolution imaging. The Bluefin-21 autonomous underwater vehicle (AUV) – which was used in the search for the lost Malaysian airliner, MH370 – used acoustic technology to map the seabed, as well as to supplement its inertial navigation sensors.
The USBL system it used emits regular acoustic signals from a beacon suspended just below the water’s surface, and this is then used to determine the drone’s position relative to the beacon. Range is calculated from the timing, and angle is determined with a small array of transducers.
USBL has great potential for underwater drones, because the transducer array is small enough to mount on the drone, allowing the vehicle itself to determine its position based only on signals from a dumb beacon.
One issue that recently came to light about active sonar systems like USBL, is the possible impact on marine life. Research has shown that military sonar used by the US Navy causes whales to breach and flee the area, affecting normal feeding and breeding behaviors. In severe cases, the animals may breach too quickly, causing decompression sickness and damaging tissues, or even causing death.
Underwater drones, which are naturally smaller than the average Naval vessel, have the advantage of using much lower powered sonar. Research shows that sound levels below 120dB will generally be ignored by whales. By keeping sonar power below this level, as well as taking precautionary measures such as gradually ramping up sonar power, so as to give early warning, the risk of impact to the marine mammals is greatly reduced.
While underwater drones lack the communication and navigational technologies we take for granted, by adapting to their environment and using the power of sound, they’re able to succeed in a vastly different environment from what we’re used to.