top of page

Our Vehicle

render.png
Çapa 1

Mechanical System

 

The Nautronics vehicle is built around a modular mechanical architecture that brings together durability, sealing reliability, and precise maneuverability. Its HDPE main frame, acrylic electronics compartment, 8-thruster propulsion system, and mission mechanisms work together to support underwater task execution. In the latest version, the manufacturing method, sealing approach, and operational workflow were improved to create a more reliable and testable system.

Hull Design and Sealing

The main frame of the vehicle is manufactured from HDPE, a material with high impact absorption capacity. This frame forms a cage-like structure that protects the central sealed unit while also allowing modular equipment mounting. The electronic components are housed inside a cylindrical compartment made of high-strength acrylic. Sealing is achieved through a low-tolerance flange, cap, and O-ring supported structure.

In the latest version, the risk of micro-voids that may occur due to the layered structure of the previous 3D printing method was reduced. With the acrylic tube compartment, a more visible, traceable, and structurally integrated solution was targeted.

Modular System and Propulsion Architecture

The vehicle moves with 8 strategically positioned thrusters. Four horizontal thrusters are placed in a 45-degree vectored configuration to improve maneuverability, while four vertical thrusters provide depth control and stabilization. This layout allows the vehicle to move in 6 degrees of freedom. The HDPE material’s density characteristics, which are close to the density of water, also support natural buoyancy and reduce the need for additional ballast.

torpido.png

Torpedo System

 

The vehicle includes two different torpedo launch mechanisms. In the timer-controlled system, the torpedo receives a start signal from the main vehicle and activates its own motor and timer circuit. In the mechanically driven system, the launch is performed through the controlled release of compressed spring energy.

Marker Dropper System

 

The electromagnetic payload release system operates without the need for motorized, multi-part mechanical arms. By reducing mechanical friction and system complexity, it provides a lighter, simpler, and faster-reacting mission mechanism.

marker.png

Grabber Mechanism

 

The grabber mechanism is one of the vehicle’s mission-oriented mechanical subsystems. Positioned within the modular body structure, it supports the need for physical interaction during underwater tasks.

 
grabber.png
_edited.jpg
WhatsApp Image 2026-05-23 at 14.43.27.jpeg

Analysis and Simulation

 

The vehicle geometry and mechanical components were evaluated through different analysis outputs. CFD visuals show the flow behavior around the body, while structural analysis outputs show displacement and stress distribution under load. Temperature analysis visuals were also used to observe the thermal distribution across the system.

 
WhatsApp Image 2026-05-18 at 17.38.26 (1).jpeg

CFD analysis visualizing the flow behavior around the projectile used in the dropper launching mechanism.

 
 
 
 
 
 
WhatsApp Image 2026-05-18 at 17.38.27 (2).jpeg

Displacement behavior of the main vehicle body under hydrostatic pressure at an approximate depth of 25 meters.

 
 
 
 
 
 
WhatsApp Image 2026-05-18 at 17.39_edited.jpg

Hydrodynamic analysis showing the flow behavior, drag, and thrust characteristics experienced by the thruster under underwater operating conditions.

 
 
 
 
 
 

Electrical System

 
 

The Nautronics AUV uses a custom electronic architecture designed for autonomous control, power distribution, and hardware safety. The system is built around an STM32-based flight controller and a custom power distribution board developed by our team. Together, these components create a fast, flexible, and direct bridge between the high-level autonomy software and the low-level hardware control layer.

mmnDiagramEnLatest.drawio.png

Custom STM32 Flight Control System (FCS)

To bypass the operational constraints of COTS solutions like Pixhawk, we engineered a custom STM32-based Flight Control System. Optimized for low-latency underwaterautonomy, this dedicated hardware handles real-time sensor fusion, multi-axis thrusterallocation, and high-speed telemetry management. The custom PCB architecture maximizesI/O flexibility within a strict footprint, acting as a direct, unbottlenecked bridge between ourhigh-level autonomy algorithms and low-level hardware execution. 

kontrolcü.png

Custom Power Distribution Board (PDB)

Engineered in house, our custom PDB forms the core electrical infrastructure of the AUV. Itfeatures multi stage voltage regulation for sensitive microelectronics and a dedicated high-current network for the thruster ESCs. Integrated with continuous overcurrent, thermal, andvoltage monitoring, the PDB ensures secure power routing and hardware protection during alloperational phases.

pdb.png

Software System

 

The software architecture of the Nautronics AUV is built on ROS 2 Humble. The system brings together mission planning, sensor data processing, vision-based perception, simulation, and vehicle control processes within a modular structure. This architecture supports rapid development with Python and performance-oriented operations with C++, while aiming to create a low-latency and testable autonomy pipeline for underwater missions.

In the latest development cycle, the software system was redesigned especially around mission planning, localization, perception, and simulation-based testing. This allows mission logic, control algorithms, and perception pipelines to be validated in a virtual environment before moving on to physical tests.

Mission Planning

Last year, our task management and mission execution relied on behaviortree_cpp. This year, we completely transitioned our pipeline to py_trees, a Python-based ROS 2 behavior tree framework. This strategic design decision was driven by our team's evolving dynamics; our newly recruited software members have a much stronger proficiency in Python. Shifting to py_trees significantly lowered the learning curve and provided a highly intuitive, easy-to-use environment for the team. It enables us to rapidly develop, test, and integrate complex mission logic and robust fallback mechanisms without sacrificing the modularity and reactivity required for autonomous underwater operations.

 

​Simulation & Hardware-in-the-Loop Testing

 

To maximize testing efficiency and minimize the risks associated with wet tests, we heavily utilize Gazebo Harmonic for simulation. We recreated the RoboSub competition elements and our AUV's URDF model within Gazebo. This environment allows us to continuously validate our controllers, test mission planner logic, and refine mechanisms like our torpedo launcher before any code touches the physical Pixhawk flight controller or Jetson board. This approach significantly speeds up our iteration cycles and ensures software readiness during hardware downtime.

WhatsApp Image 2026-05-19 at 15.18_edited.png
WhatsApp Image 2026-05-18 at 17.30.36.jpeg

Underwater Perception and Pose Estimation

 

In the underwater perception process, instead of relying only on bounding box-based object detection, YOLO Pose is used to extract semantic keypoints from competition elements. The obtained 2D image coordinates are then processed through a Perspective-n-Point algorithm to estimate the 6D pose of the target relative to the camera.

This relative pose data is directly used as an input for the Visual Servoing controller, supporting the vehicle’s real-time target tracking and alignment.

 

Control and Monitoring Interface

The software system is supported not only for running autonomous missions, but also for live vehicle monitoring and test process management. The telemetry interface allows key values such as motor status, sensor data, battery level, system temperature, and external pressure to be tracked in real time.

This interface makes it easier to observe the vehicle’s behavior during testing and detect potential error conditions more quickly.

 
PHOTO-2026-05-18-17-30-36.jpg
bottom of page