THIS IS TRAMOLA

TRAMOLA is an Unmanned Surface Vehicle (USV) project by the H-MECH Engineering society. Here is all the information about the project.

Who Are We?

Our Supporters

RoboBoat Process

TEAM TRAMOLA (LinkedIn Makalesi Kapak Görseli) (4).png

Meet The Units

Here is our units that will carry us to success. Our team consists of four units. You can scroll down for detailed information about the units.

Electronics Unit

The electronic unit of the Tramola unmanned surface vehicle is responsible for performing power calculations, power distribution, and managing the generation and packaging of the power supply necessary for the operation of the vehicle's computer, control board, and motors. Additionally, it supports the use of various sensors and facilitates communication between these sensors.

The *Jetson Nano 4GB* was selected as the onboard computer due to its image processing capabilities and high performance. For image processing, the *Dexim v28 webcam* was chosen.

As the control board, *Pixhawk* was preferred for its precision in control and navigation tasks.

For the propulsion system, *brushless motors* were selected for their high efficiency. To meet the high starting torque requirement, motors with a *920kV rating* were used.

The motors will be controlled using the *Blu30A ESC*. This ESC was chosen for its high current limit and its ability to operate bidirectionally.

Power distribution is managed via a custom-designed power distribution board created by our team. This board includes a *5V 3A output for the Jetson Nano* and a total of four main outputs. Two of these outputs are allocated for the vehicle’s motors, one is reserved for the water motor, and the final one is left as a spare.

Since Tramola consists of two torpedoes, the battery was split to maintain balance. Each torpedo contains four series-connected LiFePO4 batteries, resulting in a system with 4 series and 2 parallel connections, providing a total energy capacity of 12.8V - 12000mAh.

Software Unit

The unmanned maritime vehicle named Tramola was designed to autonomously navigate and adapt to environmental conditions. The project's main computer is a Jetson Nano, which manages all software controls for the vehicle. A Pixhawk was integrated for speed and motor control, and data from sensors like GPS and IMU were processed through the Pixhawk.

The Jetson Nano not only handled image processing but also enabled the vehicle to perform environmental analysis and make autonomous decisions. Data from the camera was processed in real-time on the Jetson Nano. ROS Melodic was used to facilitate communication between different parts of the software, enabling seamless connection between the Jetson Nano and Pixhawk via MAVROS. Additionally, the Jetson Nano was connected to a ground station via WiFi, allowing data exchange and remote control of the vehicle.

The YOLOv4 algorithm, part of the open-source deep learning framework Darknet, was utilized for image processing. Darknet was chosen as a fast and efficient object detection tool. After processing the visual data, this information was used in decision-making algorithms to achieve autonomous control of the vehicle.

NVIDIA's parallel computing platform, CUDA, and TensorRT technologies, embedded in the Jetson Nano, played a significant role in accelerating deep learning models. These technologies made it possible to run complex models like YOLOv4 in real-time on the Jetson Nano.

Mechanical Unit

Our bot has been specifically designed to overcome challenges it may encounter in the race area and to alleviate the workload of other units within the team. Our vehicle features a miniature catamaran structure, consisting of two main hulls joined by plexiglass, measuring 90 cm in length and 60 cm in width.

The catamaran design ensures stable movement of the bot against waves and wind in the race area. The surface area of the hulls in contact with water has been maximized, distributing the pressure exerted on the water and increasing stability. Additionally, fins placed at the base of the hulls reduce lateral tilt, optimize water flow, and enable higher speeds with lower energy consumption.

The design incorporates two motors positioned at the rear of the hulls. The number of motors was chosen to allow for sharp and fast maneuvers on the designated course. The motor placement was prioritized to ensure balance and efficiency during high-speed movements and maneuvers. Steering is achieved through power control.

To meet the precision and speed requirements of the production process, 3D printers were utilized. Producing the components separately and assembling them later facilitated adaptation to regional modifications, reduced waste, and positively impacted costs. PLA material was used in the production process. This material not only provides the bot with adequate buoyancy on water but also maintains its lightweight nature due to its low density.

Marketing Unit

Our primary goal as the Marketing Unit is to ensure that TRAMOLA stands out as a high-quality, recognizable, and professional brand. We achieve this through the establishment and management of social media accounts, content creation, and web design.

Social media management and content production primarily aim to promote the device and serve as a powerful advertising tool. On the other hand, our website design caters to a more professional audience by providing a platform for companies and individuals who want detailed information about the device. The website is designed to include clear, comprehensive, and precise technical specifications, ensuring easy access to critical information. This approach is particularly beneficial when participating in competitions, where presenting a thorough understanding of the device is essential.