How about a robot for exploring the sea bed? That's the sea bed and not the bottom of a paddling pool, so there would be considerable challenge in designing and making the vehicle, especially where it has to withstand reasonable pressure and go deep enough to get to the bottom of a typical harbour. It should have a steerable camera, and a way of sending video and/or telemetry back to the operator. That's a not inconsiderable challenge.
It should be steerable, of course, but it might have some intelligence of its own (it depends on the focus of your project, whether it is mechanical, electronic or computing-orientated).
Or you could assume a submersible and just focus on the command and control and telemetry and autonomous exploration, although that would miss out part of the challenge. You might need to simplify it in that way, though, for a project with a deadline.
It is easy to do this kind of thing on the land, and quite common now to do it from the air, but under water is not so easy.
Marcus