Embedded Files
Saturday 3rd June
Continuing my exploration of the 'BlueRay' project, I established the underwater robotics club at my school. I wanted to provide students with an avenue to explore the exciting world of robotics, and share my passion for the somewhat niche, but essential, underwater robotics I have been gripped with as of late. We hope to work with key stakeholders in the preservation of underwater environments such as local marine biologists and to grow our skills as innovators through competitions. Through this project, we want to foster collaboration and ignite a passion for STEM among in our school community while also raising awareness about sustainable efforts and progress in the underwater world.
Building from scratch, our team has gained invaluable skills to tackle the seemingly insurmountable task of creating a system which is both water tight and modular. Starting with a basic kit, we were able get started quickly with the assistance of a guide and start learning. The first robot, was based off the BlueDot ROV kit which we decided to modify in order to become autonomous with the exclusion of the tether. This gave us a proof of concept and quickly introduced us to the field and some of the design patterns that were common in waterproofing.
Since my last post, the design has changed significantly from the initial version. The BlueDot was hard to access requiring us to unscrew a series nuts and bolts in order to make any changes - a process that took the majority of the time available in a session to perform. We also quickly realised that we needed a larger compartment: the BlueDot had created a rigid setup and was very tight for space. The structure in the spherical container which didn't make it easy to add new circuitry or perform basic processes such as changing the battery and adjusting the wiring.The frame of the BlueDot, subject to a constant tension and not well optimised for the load it would be handling.
Having understood these issues, we decided to take what we'd learned and build on these issues. We started by switching the main structure of the robot to a tube-based design. There were a couple of reasons for doing so. Firstly, the parts it could be built out of a pre-existing water proof solution from sewer pipes, giving us the ability to buy an off-cut of a pipe as the basis to work from. Secondly advantage was the larger electronics compartment and the screw based lid which would let us have a retractable component which we could remove and work on then put back it. The main challenge then was to find a way for the wires to leave the main compartment without compromising the waterproof hull.
This was our first iteration for the tube-based model and was well thought out. The core premise of the design relies on the ability of the O-ring to separate the component into two compartments - watertight and waterlogged. Passing the wires through this connection and then filling the connection with epoxy, which hardens into plastic, allows the wires to pass through while maintaining a waterproof wall between the two sections. Though extremely promising, we found that the O-ring was quite hard to adjust in practice. In order for the O-ring to properly insulate the boundary between the two, it must be taught and moisturised to ensure that no cracks form in the rubber which could allow water through. When we tested, there was still some water which was able to penetrate the edges of the o-ring seal. This meant that the O-ring was not reliable enough and it wasn't safe enough to test with the electronics attached inside. We thus decided to overhaul the design and fill the entire cap with epoxy.
The second approach, which ended up proving successful was to use waterproof terminals and fill the entire connection of the wires to the end cap with epoxy. This way the entire cap would be water tight. Instead of bothering with unstable friction based approaches, we instead opted to use a special glue that would chemically bond the two cap and the sewer pipe together. This way the two pieces of plastic become one contigous piece and the water necessarily cannto pass through the end cap. Testing this design proved successful and eliminated much of the clutter from the previous design. Additionally, the system is more modular - the connectors are able to disconnect and change port easily.
Now that the basic robot has been built, the software and simulation aspects of the SEA-Dragon project can kick into high gear. During the summer break, I aim to develop 'Bermuda,' an open-source software framework. The software will provide a user-friendly platform for rigging, simulating, and controlling virtual autonomous underwater vehicles - fostering collaboration and accelerating progress within the underwater robotics community. Things are starting to get exciting!
Page updated
Google Sites
Report abuse