ATILIM UAV
Who we are?
Who we are?
Founded at Atılım University, the Atılım UAV Team is a passionate and multidisciplinary group of engineering students dedicated to pushing the boundaries of autonomous unmanned aerial vehicles. From structural design and aerodynamic optimization to advanced image processing, custom mapping pipelines, and rigorous system integration, we develop our aircraft almost entirely in-house. Our primary focus for the 2025-2026 academic year is competing in the prestigious SUAS competition. Here, we aim to showcase our engineering capabilities through complex autonomous flight missions, precision payload delivery, and real-time target detection under challenging field conditions.
Beyond just building a high-performance drone, our team serves as a dynamic hub for innovation, teamwork, and continuous learning. We bring together diverse talents—ranging from aerospace and mechanical engineers to software developers—to tackle real-world aviation challenges. By fostering a culture of rigorous simulation-to-field testing and creative problem-solving, the Atılım UAV Team not only prepares its members for the future of the aerospace industry but also strives to leave a lasting mark of excellence in international engineering competitions.
Founded at Atılım University, the Atılım UAV Team is a passionate and multidisciplinary group of engineering students dedicated to pushing the boundaries of autonomous unmanned aerial vehicles. From structural design and aerodynamic optimization to advanced image processing, custom mapping pipelines, and rigorous system integration, we develop our aircraft almost entirely in-house. Our primary focus for the 2025-2026 academic year is competing in the prestigious SUAS competition. Here, we aim to showcase our engineering capabilities through complex autonomous flight missions, precision payload delivery, and real-time target detection under challenging field conditions.
Beyond just building a high-performance drone, our team serves as a dynamic hub for innovation, teamwork, and continuous learning. We bring together diverse talents—ranging from aerospace and mechanical engineers to software developers—to tackle real-world aviation challenges. By fostering a culture of rigorous simulation-to-field testing and creative problem-solving, the Atılım UAV Team not only prepares its members for the future of the aerospace industry but also strives to leave a lasting mark of excellence in international engineering competitions.
ATILIM UAV TEAM — SUAS 2026
What Our Mechanics Team Has Achieved?
Requirement Analysis and Task Allocation
SUAS competition requirements were reviewed from a mechanical design perspective. Initial aircraft requirements such as frame layout, payload placement, weight limits, manufacturability, maintenance access, and structural reliability were identified. Mechanical responsibilities were divided within the subteam.
Concept Development and Design Selection
Preliminary airframe concepts were developed. Different frame layouts, arm configurations, landing gear concepts, motor mount designs, and payload integration options were evaluated. The main design direction was selected based on simplicity, strength, manufacturability, and ease of assembly.
Initial CAD Modeling and Virtual Assembly
Initial CAD models of the airframe and mechanical components were created. Motor mounts, arm connections, plate geometry, landing gear interfaces, and payload-related mounting regions were designed. Early assembly checks were performed in CAD to detect possible interference and integration problems.
Design Refinement and Optimization
The mechanical design was refined according to component dimensions and expected loading conditions. Motor, propeller, battery, electronics, and payload placement were finalized together with the relevant subteams. Critical connection points and load-carrying parts were redesigned for better strength and easier manufacturing.
Structural Analysis and Design Revision
Structural analysis was performed on critical mechanical components such as motor mounts, frame connectors, and load transfer regions. Expected thrust, motor torque, clamp forces, and safety factors were considered. Mesh refinement and design revisions were completed based on analysis results.
Prototype Manufacturing and Fit Testing
Prototype parts were manufactured and tested for fit and assembly compatibility. 3D-printed and/or machined components were inspected. Problems found during assembly, such as tolerances, hole positions, fastener access, and part alignment, were corrected in the CAD model
Final Manufacturing and System Integration
Final mechanical parts were manufactured and assembled. The frame, arms, landing gear, motor mounts, payload mounts, and other structural components were integrated. The mechanical subteam supported full aircraft integration with avionics, software, and payload systems.
Ground Testing and Mechanical Inspections
Ground tests and pre-flight inspections were supported. The team checked fasteners, frame stiffness, motor mount security, vibration behavior, landing gear strength, cable clearance, and general mechanical reliability. Required improvements were applied after each test.
Flight Testing and Maintenance Procedures
Flight test support was provided. After each flight, the aircraft was inspected for loosening, cracks, deformation, vibration-related problems, and alignment issues. Final reinforcements, spare parts, maintenance procedures, and competition field tools were prepared.
Competition Preparation and Field Support
Final competition preparation was completed. The aircraft was mechanically inspected before shipment and before flight operations. Spare mechanical components, tools, fasteners, and repair materials were organized. The mechanical subteam provided field support during the competition.
What Our Electronics & Communications Team Has Achieved?
System Architecture and Component Selection
The SUAS 2026 competition requirements were analyzed to define the aircraft's avionics system architecture. System responsibilities were assigned, communication interfaces were identified, and the preliminary selection of avionics components was completed.
Power Budget and Hardware Selection
A power budget was established based on the estimated power consumption of the main avionics’ components. Appropriate batteries, DC-DC converters, and electrical connectors were then selected to meet the system requirements.
Communication Network and Procurement
The avionics communication network was finalized by defining the interfaces and communication protocols between the major subsystems. The overall electrical architecture was completed, and procurement of electronic components was carried out.
Platform Assembly and Telemetry Verification
The power distribution system was assembled and validated through initial electrical tests. Long-range telemetry communication and MAVLink integration were successfully verified using Mission Planner. By the end of the month, the first fully functional avionics platform had been completed.
Airdrop Mechanism and Payload Testing
The prototype of the airdrop subsystem was designed and assembled. Suitable DC motors, servos, and motor drivers were integrated into the release mechanism. A series of payload release tests were conducted from different altitudes to evaluate deployment reliability and mechanism performance.
Initial Flight Tests and Performance Monitoring
The avionics safety mechanisms were integrated into the system, and the first flight tests were successfully conducted. During the flight tests, the aircraft successfully followed predefined waypoints uploaded through Mission Planner, validating the navigation and flight control systems. Power consumption, current draw under load, voltage stability, and battery performance were monitored to evaluate the electrical system under real operating conditions. Necessary improvements were implemented based on the collected test data.
System Optimization and Reliability Testing
The avionics system was optimized based on the results of previous flight tests. Electrical connections, cable routing, and power distribution were improved to increase system reliability. Repeated flight tests were conducted to validate telemetry performance, payload operation, and overall avionics stability.
Autonomous Mission Validation
Autonomous mission functions were extensively validated through repeated flight tests. Waypoint navigation, payload release timing, telemetry communication, and fail-safe operations were verified under different mission scenarios to improve mission reliability.
Mission Simulations and Endurance Testing
Full mission simulations were conducted under competition-like conditions. Long-duration flight tests were performed to evaluate battery endurance, thermal performance, communication stability, and the reliability of the integrated avionics system. Final hardware adjustments were completed based on the test results.
Competition Preparation and Technical Support
Final pre-competition inspections and system verification were completed. Spare avionics modules, batteries, connectors, and repair equipment were prepared. The avionics team provided system maintenance, troubleshooting, and technical support throughout the competition.
What Our Software Team Has Achieved?
Requirements Analysis and Task Allocation
SUAS competition requirements were reviewed from a mechanical design perspective. Initial aircraft requirements such as frame layout, payload placement, weight limits, manufacturability, maintenance access, and structural reliability were identified. Mechanical responsibilities were divided within the subteam.
Software Architecture Design and Infrastructure Selection
Preliminary software architecture concepts were developed. Onboard versus ground-based inference, candidate detection model families, and mapping approaches such as full photogrammetry versus lightweight incremental stitching were evaluated. The main design direction was selected — onboard detection on an embedded GPU computer, Mission Planner based mission control, and an in-house mapping pipeline — based on simplicity, control, and field reliability.
Environment Setup and Initial Data Collection
Development environments were set up for the Jetson companion computer, cloud-based model training, and dataset management. Initial aerial datasets were collected and labeled, and the camera video link (RTSP) was integrated and tested. Early object detection experiments were performed on recorded footage.
Initial Model Training and Ground Control Integration
The first detection models were trained and evaluated. The ground control flow between Mission Planner, the companion computer, and the camera was defined, and detection outputs were connected to the mission logic. Initial performance tests revealed the operating limits of the first models and guided the data collection plan.
Comprehensive Simulation Setup and Virtual Validation
A comprehensive simulation environment integrating Gazebo, SITL, ROS 2, ArduPilot, and pymavlink was established. The environment was built as a distributed architecture mirroring the field setup, with the Jetson companion computer connected as a separate networked node and camera imagery streamed over RTSP. Autonomous flight scenarios and mission logic were validated virtually before field testing.
Custom Mapping Pipeline (FastMosaic) Development
The in-house mapping pipeline, FastMosaic, was developed. GPS-assisted frame selection, feature-based stitching, and blending stages were implemented and tested on recorded flight imagery, producing map outputs directly in the competition submission format. Camera stream test utilities were added for pre-flight verification.
Autonomous Guidance Integration and Failsafe Testing
Detection-to-guidance integration was completed so that confirmed detections could be handed to the autopilot for autonomous positioning over the target. Failsafe scenarios such as GPS degradation, link loss, and low battery were exercised in simulation. The dataset was expanded with additional aerial imagery.
Field Data Collection and Model Performance Analysis
Field data collection flights were performed and the resulting aerial imagery was added to the dataset. Evaluation on real flight footage exposed the weaknesses of the initial model, particularly the loss of detections at distance. A systematic diagnosis of the data and training configuration was started.
Root Cause Analysis, Dataset Optimization, and Retraining
The detection failure was root-caused to missing small-object examples and disabled scale augmentation. A consolidated dataset of approximately 16,000 images was built through merging, deduplication, and class rebalancing, and the model was retrained at higher resolution, reaching mAP50 0.95 on an aerial-only validation set. Tiled inference was studied and profiled on the Jetson, and the final real-time flight configuration was selected and deployed.
End-to-End Field Testing and Final Competition Preparation
End-to-end field testing was completed: both targets were detected from approximately 20 m and payloads were delivered autonomously, with additional low-light detection trials performed. Software sections of the Technical Design Report were written. Final competition preparation was carried out, including configuration freeze, operating threshold checks, and field operation procedures.
