Engineers at MIT have built a wireless, battery-free underwater camera capable of collecting its own energy while consuming very little power, according to a new paper published in the journal Nature Communications.
The system can take colour photographs of remote underwater objects – even in dark environments – and will transmit data wirelessly to monitor the underwater environment in real-time, help discover new rare species or monitor ocean currents, pollution or commercial and military operations.
A wireless underwater camera without batteries
We already have a variety of methods for taking underwater images, but according to the researchers, “most marine and oceanic life has not yet been observed. This is because most existing methods of taking underwater images require a connection to a ship, underwater drone or power plant for power and communication.”
Those methods that do not use network sharing must contain battery power, which limits their lifespan. While it is in principle possible to harvest energy from waves, underwater currents and even sunlight, adding the necessary equipment would result in bulkier, more expensive underwater cameras.
The team at MIT, therefore, set out to develop a solution for a battery-free wireless imaging method. The design goal was to reduce the amount of hardware required as much as possible. They also wanted to keep power consumption to a minimum, so they used an inexpensive off-the-shelf imaging sensor.
But this sensor only produces grey-scale images. The team also needed to develop a low-power flash, as most underwater environments don’t have much natural light.
An overview of how the underwater backscatter imaging system works
Both solutions to the challenge proved to combine red, green and blue LEDs. the camera uses the red LED for in-situ illumination and captures that image with its sensor, then repeats the process using the green and blue LEDs.
According to the authors, the image may look black and white, but the light from the three colours of the LEDs is reflected in the white part of each image. As a result, it is possible to reconstruct the full-colour image in post-processing.
“When we took art classes as children, our teachers had taught us that we could use the three basic colours to make all the colours,” says co-author Fadel Adib. “For the colour images we see on the computer, the same rules are followed: we only need three channels – red, green and blue – to build a colour image. “
After the image data has been encoded as bits, the sensor relies on piezo-acoustic backscatter for ultra-low-power communication, rather than batteries. Instead of generating its own acoustic signal (e.g. sonar), this method relies on modulating the reflection of incident underwater sound to transmit one piece of data at a time.
This data is picked up by a remote receiver capable of recovering the modulation pattern, which then uses binary information to reconstruct the image. The researchers estimate that their underwater camera is about 100,000 times more energy efficient than comparable cameras and can operate continuously for several weeks.
Of course, the team built a proof-of-concept prototype and conducted a number of tests to prove that their approach works. For example, they imaged pollution (in the form of plastic bottles) at Keser Pond in southeastern New Hampshire, and imaged an African starfish (Protoistoisto Linkley) in a “controlled environment with external lighting”. The latter image is of sufficient resolution to capture the various nodules on the five arms of the starfish.
Sample images obtained using underwater backscatter imaging
The team was also able to use their underwater wireless camera to monitor the growth of an aquatic plant (Aponogeton Ulvaeus) over a period of several days and to detect and locate visual tags normally used for underwater tracking and robotic operations. The camera achieved high detection rates and high localisation accuracy at a distance of approximately 3.5 metres (approximately 11½ feet). The researchers believe that longer detection ranges can be achieved using higher-resolution sensors. Distance is also a factor in the camera’s energy harvesting and communication capabilities, according to tests conducted on the Charles River in eastern Massachusetts. As expected, both of these key capabilities decreased with distance, although the camera successfully transmitted data up to 40 metres (131 feet) from the receiver.
In short, “the untethered, inexpensive and fully integrated nature of our approach makes it ideal for large-scale marine deployments,” the authors write. Scaling up their approach would require more sophisticated and efficient transducers, as well as higher-powered underwater acoustic transmissions. One could also use existing mesh networks of buoys on the ocean surface or underwater robotic networks such as Argo buoys to provide remote power for energy harvesting cameras.
“Personally, one of the most exciting applications for this camera is in the context of climate monitoring,” says Adib. “We’re building climate models, but we’re missing data from more than 95% of the ocean. This technology can help us build more accurate climate models and better understand how climate change is affecting the underwater world.
