In 2018, together with my teammate İlayda Dülüloğlu, we developed the initial vision of SUFAI and officially founded it. As reflected in its full name — SUFAI (Space, Unmanned Vehicle Systems, Future, Artificial Intelligence) — the initiative aims to conduct advanced R&D primarily in unmanned vehicle systems, as well as in the fields of space, artificial intelligence, and future technologies, leveraging cutting-edge innovations to create valuable contributions for humanity.

With the motto “The Future Beneath Our Wings”, SUFAI has successfully completed dozens of projects, won numerous international awards, and broken records. Our work continues to this day. İlayda and I are still actively managing the initiative.

I designed an unmanned aerial vehicle (UAV) from scratch for the Teknofest International Combat UAV competition. I successfully produced the prototype and completed its flight tests. Featuring a fixed-wing structure with electric propulsion, this was one of the first UAV systems I personally developed.

After numerous design iterations and failed flight tests, this project helped me gain hands-on experience with many subsystems found in mini-class UAVs. It provided significant insight into the areas I needed to focus on in the future.

I also continued to use this platform as a reference point in the following years and tried to improve it further. This accumulated experience enabled us to move quickly during the conceptual design and prototyping phases of later UAV projects.

To demonstrate the performance of STALKER, a fixed-wing unmanned aerial vehicle I designed and manufactured for the Teknofest International Combat UAV competition, I conducted a long-range flight with real-time video transmission and telemetry data. This flight led us to break the national inter-university record in Turkey.

Following this achievement, and in response to requests from other universities in the Aegean region, my team and I conducted presentations and delivered in-person training sessions on long-range mini-class UAV systems at Dokuz Eylul University, Ege University, Yasar University, and Izmir University of Economics.

During the COVID-19 pandemic, several precautions and measures were implemented worldwide. I presented one of my project ideas on this topic to Prof. Dr. Nükhet Hotar, the then-rector of Dokuz Eylul University, together with my teammate İlayda Dülüloğlu, and she requested that the project be brought to life.

Working quickly, I developed a fully functional TSH-6 level prototype in just two weeks. ETRAS was capable of performing multiple functions in about 2 seconds, including facial recognition via image processing, mask detection with a custom-trained deep learning model, HES code verification, and temperature measurement.

With support from Dokuz Eylul University’s DEPARK technology office, the patent process was successfully completed.

Until that year, there was no laboratory, workshop, or even an academic at Dokuz Eylul University working on fixed-wing unmanned aerial vehicles or unmanned systems.

With the support of the then-rector Prof. Dr. Nükhet Hotar, who closely followed our projects, the establishment of a Space and Aviation Laboratory for SUFAI within the Faculty of Engineering at Dokuz Eylul University was approved, followed later by recognition as a research center.

My teammate İlayda Dülüloğlu and I were solely responsible for all aspects of the workshop: architectural planning, engineering drawings, structural calculations, construction, equipment selection, installation of smart laboratory systems, and launching it into service. This laboratory is still actively producing projects today.

Over the course of approximately 18 months, I led the entire development process of the SAFIR fixed-wing unmanned aerial vehicle, including its CAD design, CFD analyses, avionics and electronics design, payload mechanisms, ground control station, autonomous flight software, autonomous target detection algorithms, ground testing, and flight testing.

To demonstrate the vehicle’s performance, I conducted a long-range flight with live video and telemetry transmission. This flight matched our previous record and allowed us to break the Inter-University Turkey Record for the second time. Additionally, we participated in the International UAV Competition with this platform and won the TÜBİTAK Special Award. (No other team or project received an official ranking award in that competition.)

In the first version of the SAFIR fixed-wing unmanned aerial vehicle, I had previously led the entire process for about 18 months — including CAD design, CFD analyses, avionics and electronics systems design, payload mechanism development, ground control station setup, autonomous flight software development, autonomous target detection algorithms, ground tests, and flight tests. In the second version of the aircraft, several additional optimizations were implemented.

During these optimizations, I used a custom-developed parametric software that works in coordination with SolidWorks and Ansys, primarily based on C and Python. As a result of these improvements, we conducted a long-range flight with live video and telemetry transmission to demonstrate performance. This flight not only matched our previous record but also allowed us to break the Inter-University Turkey Record for the third time and enter the World Record rankings.

We also participated in the International UAV Competition with this aircraft, where we broke the fastest course record in competition history by performing all maneuvers completely autonomously. Additionally, in line with the competition’s concept, this became the first UAV project in competition history to autonomously detect a fire zone using image processing and autonomously deliver its payload. (No other team or project received an official ranking award in that competition.)

I began working on this project in 2018 and completed its first version in 2021. The software was capable of autonomously detecting and tracking other aerial vehicles through image processing.

To train the model responsible for recognizing UAVs, hundreds of formation flights were conducted, and thousands of images collected from those flights were used. The core of the software was built on YOLO and GOTURN frameworks. Thanks to custom algorithms I developed, the software could process and track images at 150–100 frames per second even on low-power hardware suitable for mini-class UAVs.

This software became the first ever tested and successfully locked-on system in full compliance with target detection and tracking rules at the International Teknofest Combat UAV competition. Later on, I made the software open-source and created a complete training program to share it for public benefit.

Using the infrastructure developed in my previous projects, I designed and manufactured a fixed-wing unmanned aerial vehicle for the International Teknofest Combat UAV competition. This aircraft was also equipped with an autonomous target detection and tracking software.

Thanks to a long-range flight conducted to demonstrate the performance of this UAV, we broke the Inter-University Turkey Record for the 4th time and entered the World Record rankings for the 2nd time. Despite experiencing four crashes due to signal interference caused by protocol jamming devices and broadcast equipment present at the competition site, we built four entirely new aircraft from scratch and became the only team to participate in all stages of the competition.

Additionally, our target detection and tracking software was the first to send a lock-on package to the competition server. According to the officially published rulebook, no team was deemed eligible for an award in that year’s competition.

This project integrated artificial intelligence into the CAD design and performance analysis processes of fixed-wing unmanned aerial vehicles (UAVs). AI algorithms were used for parametric design optimization, enabling maximum aerodynamic efficiency. The AI-based parametric modeling software developed within the scope of the project worked in integration with tools like SolidWorks and Ansys.

Optimizations focused on high speed, low fuel consumption, and durability were supported by advanced Computational Fluid Dynamics (CFD) analyses. The results reduced aerodynamic drag while improving maneuverability. This project showed me how powerful the synergy between AI and engineering software can be in UAV design processes, and it allowed me to gain significant time advantages in conceptual design and prototyping phases in my subsequent projects.

This project, carried out for the Turkish Space Agency, aimed to establish a comprehensive Space Campus to strengthen Turkey’s space vision. The campus consists of state-of-the-art R&D facilities for space technology development, astronaut training, and rocket systems research.

One of the most striking phases of the project was the full-scale modeling and prototyping of the Moon’s surface. This model was designed as a testing and training ground for space mission simulations. Additionally, as part of the project, a scale model of the campus—which aligns with the strategic goals of the Turkish Space Agency—was presented to President Recep Tayyip Erdoğan with the support of then-Rector of Dokuz Eylul University, Prof. Dr. Nükhet Hotar. The model received great attention as a symbolic representation of Turkey’s commitment to space exploration and its international ambitions.

This project was a prototype-level UAV development effort inspired by the high-speed and autonomous capabilities of the SAFIR UAV, designed for use in kamikaze missions. The UAV’s main structural design was optimized to ensure low weight and aerodynamic efficiency, allowing it to move rapidly while maintaining a lightweight structure.

During production, composite materials manufactured via additive manufacturing were used to ensure minimal weight and maximum durability. The kamikaze UAV developed within this project was equipped with image processing software optimized for autonomous target detection and neutralization. The UAV was capable of processing image data from its sensors in real-time and locking onto both stationary and moving targets.

A ROS (Robot Operating System)-based flight software, derived from previous projects, demonstrated reliable performance in autonomous route planning and mission execution. The autonomous target detection software used in this project was supported by advanced image processing algorithms. Visual data from optical sensors was analyzed by an AI model optimized for low computational power. The algorithm used in the prototype determined and verified targets, then calculated the ideal trajectory for the fastest possible strike. This process demonstrated great success in both speed and accuracy.

The project was completed at the prototype stage, and a modular architecture was adopted during the design phase to allow for future improvements and integrations.

This project aimed to develop a rocket system capable of vertical landing. Conducted in collaboration with the defense industry, it was carried out through theoretical and engineering phases and was completed at the project stage without producing a fully functional prototype. The rocket was designed with a thrust system based on advanced control algorithms to enable vertical descent. The focus was on the synchronized operation of multiple thrusters and autonomous landing algorithms to ensure stability during descent.

The landing process was optimized using PID-based control systems, aiming to properly direct the thrust of the motors. The landing algorithm developed as part of the project included a system design that would allow the rocket to reduce its speed during descent and make a controlled soft landing. In addition, data from sensors used during descent would be processed in real time to determine the rocket’s position and velocity, ensuring a safe landing.

This project was completed through engineering analysis and simulation studies without advancing to the fully functional prototype phase, establishing a solid foundation for future development and testing stages.

The Haris VTOL UAV v1 Project was a mini unmanned aerial vehicle developed by SUFAI prior to its formal incorporation, designed for reconnaissance, surveillance, and light-capacity logistics transport. In this project, a fixed-wing UAV was equipped with vertical takeoff and landing (VTOL) capability to create a versatile platform. Powered by a 4+1 electric propulsion system, the UAV delivered seamless performance in both vertical takeoff/landing and forward flight modes.

In the design of the Haris VTOL UAV, vertical takeoff and landing were stabilized through an electric propulsion system, while the fixed-wing configuration provided high efficiency during forward flight. The UAV was capable of carrying small-scale logistics payloads and was optimized for effective use in surveillance and reconnaissance missions. During the prototyping phase, it successfully completed missions such as vertical takeoff, forward flight, and vertical landing. This project served as a proof of SUFAI’s engineering capabilities before incorporation and laid the foundation for larger autonomous systems projects in the future.

This project aimed to leverage our autonomous control experience gained from UAVs to develop the necessary hardware and software solutions to transform a manually operated sailboat into an autonomous vessel. Within the framework of a collaboration with MİLRES, the design of autonomous control systems, a ground control station interface, and communication systems was completed and delivered to the company as a project proposal.

As part of the project, the necessary sensor integrations and control algorithms were developed to enable the sailboat to autonomously determine its route and adjust its sails based on wind direction and speed. In addition, the autonomous software developed for route optimization and maneuverability was designed to determine and execute the most efficient sailing path under dynamic sea conditions. The software was also capable of detecting other surface elements and autonomously re-planning its course to avoid collisions.

The user-friendly ground control station interface designed for communication with the sailboat was developed to provide real-time control and monitoring capabilities. The communication infrastructure was planned to ensure stable long-range data transmission. The project, including its hardware and software details, was submitted to the company but did not proceed to the prototyping stage. This project formed a strong foundation for future innovations in surface autonomous systems that I plan to develop.

SAFDER Combat UAV v3 was developed as a fixed-wing UAV built upon the successes of previous versions, with a focus on optimizing autonomous target detection, tracking, and kamikaze systems. Throughout 2022, this version proved its performance in various tests and competition events, achieving a series of milestones including a world record ranking and notable competition results.

The algorithms enabling fast decision-making by the autonomous systems during flight were restructured. The YOLO and GOTURN-based image processing framework was made more efficient in terms of computational load. Additionally, the PID-based flight control system was re-optimized to ensure stability at high speeds. At the Samsun Teknofest International Combat UAV Competition, SAFDER UAV v3 became the first and only project to successfully lock onto all competing UAVs thanks to its enhanced and optimized target detection software.

Unlike its predecessors, SAFDER UAV v3 was equipped with an autonomous kamikaze dive module following target lock-on. After locking onto a target, it executed the first successful autonomous kamikaze mission by diving at high speed using an improved route planning algorithm, becoming the first team to score points in that category.

The Mind-Controlled Drone Project, developed by SUFAI using advanced technologies, aims to enable unmanned aerial vehicles to be controlled directly through brain and muscle activity without any physical control devices. This project offers a solution that requires no piloting skills, allowing users to control the drone effortlessly—even during complex missions.

The system developed for this project is equipped with sensors that read muscle movements and brain signals. These sensors detect the user’s neurological and muscular inputs to guide the drone’s movement. Additionally, a smart glasses system that tracks pupil movement and gaze direction allows users to mark objects they look at, enabling the drone to focus on those targets. The human-machine interface I developed at SUFAI establishes a seamless connection between the drone and the ground control station, enabling the operator to control the drone without any physical controller—using only muscle movement or voice commands.

The drone can also autonomously detect and map surrounding obstacles and objects, providing the operator with visual, auditory, and haptic feedback. The image processing software developed for the project allows the drone to identify target elements in its environment and focus on them. Furthermore, the system can monitor the operator’s health status and either patrol the surrounding area or remain in passive follow mode.

This project represents an innovative step toward controlling aerial vehicles via cognitive and muscular inputs, once again showcasing SUFAI’s expertise in artificial intelligence and autonomous systems. The project has been successfully demonstrated to investors on multiple occasions.

Following the ongoing support and sponsorship from SANKO Holding and Chairman of the Board Mr. Adil since 2020, a project was launched to develop a Mini VTOL-class unmanned aerial vehicle with vertical takeoff and landing capability, as requested.

Through a joint stock company affiliated with the holding, I worked for over a year as a project lead together with my teammate İlayda Dülüloğlu.

Within the scope of the project, two UAVs were manufactured using 3D printing technologies and composite materials, each with a wingspan of approximately 3 meters. The main electric motor for these UAVs was personally designed by me and produced by a domestic company.

With the first UAV, we participated in the Teknofest Festival as a company and delivered presentations to various military and civilian institutions in Turkey for potential sales. Some institutions requested demo flights.

Additionally, two scaled-down VTOL UAV versions of the main vehicles were developed within the scope of the project. These scaled models successfully completed demo flights demonstrating their intended functions. After the completion of the project, all prototypes were delivered to the investor.

This project is a fixed-wing VTOL UAV initiative developed upon the request of a private company. The conceptual and detailed design phases have been completed. It is intended to be marketed primarily to countries with coastlines along open seas, including Morocco, Algeria, Tunisia, Italy, Spain, Portugal, Greece, Malta, Egypt, and Libya. Since the project is still ongoing and subject to confidentiality, more detailed information will be added at a later stage.