An underwater vehicle that caters autonomous navigation, visual identification and collection of objects via a mechanical Gripper. It was designed for Singapore AUV Challenge ‘23 (SAUVC) but its long term goal is to explore uncharted waters where humans can't reach and conduct effective research.

More about SAUVC

Held annually in Singapore, with participants all over the world, the competition consists of 2 Pre-requisite rounds — Video Submission & Qualification, and Final Round consisting of four tasks.

1. Navigation
2. Target Acquisition
3. Target Reacquisition
4. Communication & Localization

Each task has discrete conditions for its successful completion and awarded points if done right, or appropriate penalties were imposed for any misfunction.

In order to complete the tasks, the vehicle must be able to real time navigate through mildly turbulent water, identify various objects and their colors and be able to pick them up with the help of a gripper. Initially, the numerous tasks were quite overwhelming for us due to our lack of experience in constructing an underwater vehicle. This was also the first underwater vehicle developed at IIT Roorkee. Hence, we began from the ground up and gradually worked our way forward

Research and Reasoning

Whilst in the research team, I actively contributed in the following:

1. Floating and Weight Balance: The vehicle was intended to float on water when not in operation. Thus based on the dimensional constraints mentioned for the competition, I concluded the ideal weight our vehicle close to 11 kg and used filler weights for weight balance during different iterations of design.
2. Brainstorming on Vehicle Design: Since the vehicle had to be water proof in order to entail electronic equipment, we divided the design into two key-frames — a Cylindrical Hull with interior cavity of all necessary electronic equipment and Metallic frame that support thrusters and exterior attachments.
3. Thruster Positioning: I worked on providing 6 degrees of freedom so that the vehicle and reorient itself while maneuvering in mildly turbulent water and this was achievable with 6 thrusters. I tried various thruster layouts and based on requirements select the best suitable orientation. Later we reduced the degree of freedom to five, as the navigation code coupled with gyroscope sensor was not able to perform smooth maneuver and increased response time in vehicle operation.

As we were able to figure out various aspects of vehicle we headed to 3D model it in SolidWorks and develop prototypes using appropriate manufacturing processes.

CAD Modelling and Prototypes

Manufacturing Processes: Laser Metal Cutting, 3D printing, Acrylic Cutting Optimization Software: Ansys Fluent, Ansys Static Structural analysis

I proceeded with Computational Fluid Dynamics in Ansys Fluent to reduce the drag force and optimize the movement in water which also helped us determine the width of frame and optimize it’s shape. After several test runs and countless iterations, we finalized the design and Integrated all the necessary sensors required for navigation of the vehicle.

Final Model in Action