MIT is creating underwater robots that think for themselves

A research initiative at MIT is working on unmanned robots that will be able to work on their own underwater.

The robot crew is to be given a high-level goal, and then the robots would work together to determine how to best achieve this objective, according to a Discovery News report.

The approach is modeled after the starship Enterprise from Star Trek, where one robot is the captain making the highest-level decisions, followed by other robots that serve functions such as navigator, engineer, or even a repair bot.

Brian Williams, the principal developer of the mission-planning system at MIT, said that the team wanted to create vehicles that could execute missions and adapt on their own without the aid of humans. The system would have choices, like to either go to a location and map it out or communicate via a satellite link, for example.

Williams developed a similar system to this when he was working for NASA in the 1990s. So far, the MIT team has tested one of these underwater systems off the coast of Australia and will make a presentation on the system in June at a conference in Israel.

If his team can master it, it would be a huge boon for those who conduct underwater research. The underwater environment is a difficult one for human beings, and humans have relied on robots for a while to help them monitor the underwater world. However, until now robots can only do what they are programmed to do, and engineers must spend lots of time getting them to do simple tasks — which is why this research could prove to be groundbreaking.