M.Sc. Researcher · Akron, OH

Ardacan Yildiz Electrical & Computer Engineer

I build intelligent systems where deep learning meets power hardware — from EV battery estimation to real-time embedded control. Published at IEEE ECCE, with experience across defense, aerospace, and energy.

Top 0.6%
Univ. Entrance / 2.4M
5+
Engineering Roles
IEEE
ECCE 2025 Published
Ardacan Yildiz
⚡ Deep Learning
🛰 Embedded
⚙ Power Electronics

Background

I'm a graduate researcher at the University of Akron pursuing an M.Sc. in Electrical & Computer Engineering. My work sits at the intersection of artificial intelligence and power electronics — applying deep neural networks to real-world energy systems and electric vehicle battery management.

Before grad school, I developed AI-driven ADAS systems at ASELSAN, Turkey's largest defense electronics company, and built flight-critical avionics at Turkish Aerospace Industries for UAV platforms including ANKA and AKSUNGUR.

I'm drawn to problems that demand both algorithmic precision and hardware intuition — where a CNN-LSTM model meets a CAN bus, or a sensor fusion pipeline runs in real time on a DSP.

🧠

Deep Learning

PyTorch, TensorFlow, CNN-LSTM, GANs, self-attention

Power Electronics

FOC, DAB converters, SVPWM, PMSM control

🛰

Embedded Systems

TI DSP, FPGA, CAN protocols, real-time control

📊

Signal Processing

Kalman filtering, anomaly detection, INS fusion


Professional history

Defense, aerospace, and graduate research — systems built for the real world.

Graduate Research Assistant
Aug 2024 — Present
University of Akron — Deep Learning, Power Electronics & Control
  • Developing EV battery SoC estimation using task-specific deep neural networks
  • Researching DAB Triple-Phase-Shift modulation with boosting algorithms and CNN-LSTM models
AI Engineer — ADAS Project
Oct 2023 — May 2024
ASELSAN A.S.
  • Built INS Sensor Fusion and ML-based anomaly detection for driving conditions
  • Implemented IMU + GNSS sensor fusion via Kalman filtering
  • Applied autoencoders, GANs, LightGBM, and XGBoost for ADAS cavity classification
Embedded Software Developer
Aug — Sep 2023
ASELSAN A.S.
  • Developed Thermal Camera GUI and control systems in C/C#
  • Supported military-grade testing and validation procedures
Co-Op Engineer — Sky Experience
Aug 2022 — May 2023
TUSAS — Turkish Aerospace Industries
  • Developed and tested alternator circuitry for ANKA & AKSUNGUR UAV platforms
  • Contributed to FADEC development and FPGA-based landing gear driver design
  • Worked on RC decoy jet aircraft systems
Digital Electronics Intern
Jul — Aug 2022
TCDD — Turkish State Railways Corp.
  • Managed Fiber Optic Communication and GSM-R systems
  • Worked with DWDM telecom, TCP/IP, and SDH signaling

Engineering work

Hands-on projects spanning ML, power electronics, and embedded systems.

TI DSP 3-Phase Inv. CAN Vector CANalyzer Python GUI Telemetry PMSM Motor REAL-TIME MONITORING

CAN Interfacer (CAVES) — Inverter Communication

Aug 2024 — Present · AAM Sponsored

CAN bus interface between a TI DSP 3-phase inverter and host PC with live Python GUI for real-time telemetry, fault states, and PMSM angle offset calibration.

CAN BusPythonDSPCANalyzerPMSM
i_abc 3-Phase Clarke αβ Park dq PI d & q axis SVPWM Inverter FEEDBACK LOOP FIELD ORIENTED CONTROL PIPELINE

FOC Current-Control 3-Phase Inverter Drive

Aug 2024 — Present · OAI&IV Sponsored

Implemented FOC on DSP with Clarke/Park transforms and SVPWM generation. Designed and tuned PI controllers for d-axis and q-axis current loops with oscilloscope validation.

FOCSVPWMPI ControlDSPPMSM
Battery V, I, T CNN LSTM Self-Attn Mechanism SoC % Real-Time CNN-LSTM + SELF-ATTENTION ARCHITECTURE

Real-Time SOC Estimation — Deep Neural Networks

Dec 2024 — May 2025 · Published at IEEE ECCE

Task-specific CNN-LSTM architecture with self-attention for real-time battery State-of-Charge estimation. Published at IEEE ECCE 2025, Philadelphia.

CNN-LSTMSelf-AttentionPyTorchBMS
Image ResNet50 CNN YOLOv5 Ensemble Analysis Piece ID + Board Position MULTI-MODEL CLASSIFICATION PIPELINE

Chess Piece Identifier & Position Analyzer AI

Jan 2024 — May 2024

Computer vision system with ResNet50, CNN, and YOLOv5 for identifying chess pieces and analyzing board positions from images.

ResNet50YOLOv5CNNComputer Vision
Dataset Features BMI, Glucose... k-NN Log. Reg. Rand. Forest AdaBoost Ensemble Predict Diabetes ADAPTIVE BOOSTING CLASSIFICATION

Diabetes Prediction — Adaptive Boosting ML

Oct 2023 — Dec 2023

Implemented clinical diabetes prediction using k-NN, Logistic Regression, and Random Forest classifiers with AdaBoost for robust ensemble classification.

AdaBoostRandom Forestk-NNScikit-learn

Technical toolkit

Programming

PythonPyTorch TensorFlowNumPy PandasScikit-learn MatplotlibSciPy CC++ C#AssemblyVHDL

Embedded & Hardware

MATLAB/SimulinkCode Composer Studio LTspiceProteus Vector CANalyzerOscilloscope

Tools & Workflow

Git / GitHubLaTeX Microsoft OfficeGoogle Workspace

Core Domains

Machine LearningSignal Processing Power ElectronicsEmbedded Systems Control SystemsCAN Bus Protocols

Languages

Turkish (Native)English (Professional) German (Elementary)

Research

IEEE ECCE 2025 — Philadelphia, PA

Real-Time SOC Estimation Using Task-Specific Deep Neural Networks for Battery Management

A. Siddiquee, A. Yildiz, A. Uzum, S. I. Hasan, Y. Sozer, and M. J. Kisacikoglu

pp. 1–7, 2025.

10.1109/ECCE58356.2025.11260091

Academic background

M.Sc. Electrical & Computer Engineering
University of Akron, Ohio
GPA: 3.65 / 4.00 Aug 2024 — Present
B.S. Electrical & Electronics Engineering
Bilkent University, Ankara, Turkey
GPA: 2.87 / 4.00 Sep 2020 — June 2024

Let's connect

Whether you're looking for an engineer, a research collaborator, or want to talk deep learning and power electronics — reach out.