Reinforcement Learning & Autonomous Systems

This line studies decision-making under uncertainty for autonomous systems. My doctoral work at Istanbul Technical University introduced Guided Soft Actor-Critic (GSAC), a guided deep reinforcement learning method for partially observable Markov decision processes (IEEE Access, 2021).

Recent work extends GSAC to belief-aware privileged distillation for autonomous driving under partial observability (CVPR 2026 Workshop on Autonomous Driving) and to agent-based reinforcement learning for geochemical control in geothermal systems. Applied work includes cloud-assisted connected and automated mobility in the EU H2020 5G-MOBIX project (IEEE IV 2022), human-interactive social robot navigation (ICRA 2024 workshop) and reinforcement learning for COVID-19 mitigation policies.

Publications
Mehmet Haklıdır (2026). Consent Chain Degradation in Embodied Multi-Agent Systems: Bridging the Gap Between AI Agent Governance and Robot Ethics. ICRA 2026 WoRoBet.
Preprint
Mehmet Haklıdır (2026). When Does Adaptive Guidance Help? Belief-Aware Privileged Distillation for Autonomous Driving Under Partial Observability. CVPR 2026 WAD.
Preprint
Füsun Servin Tut Haklıdır, Mehmet Haklıdır (2026). Agent-Based Reinforcement Learning for Geochemical Control in Geothermal Energy Systems. GSL AI in the Geosciences.
Tuba Girgin, Emre Girgin, Yiğit Yıldırım, Emre Uğur, Mehmet Haklıdır (2024). Bidirectional Human Interactive AI Framework for Social Robot Navigation. ICRA 2024 RTSS.
Preprint
Sencer Melih Deniz, Hamraz Javaheri, Juan Felipe Vargas, Doğan Urgun, Fariza Sabit, Mahmut Tok, Mehmet Haklıdır, Bo Zhou, Paul Lukowicz (2022). Prediction of Lifted Weight Category Using EEG Equipped Headgear. IEEE-EMBS BHI 2022, 1-7.
Tahir Sarı, Mert Sever, Arda Taha Candan, Gülsüm Tuba Çibuk Girgin, Emre Girgin, Mehmet Haklıdır (2022). Cloud Assisted Connected and Automated Mobility System Architecture Design and Experimental Verification: The 5G-MOBIX Autonomous Truck Routing Use Case. IEEE IV 2022, 1374-1380.
PDF
Mehmet Haklıdır, Hakan Temeltaş (2022). Autonomous Driving Systems for Decision-Making Under Uncertainty Using Deep Reinforcement Learning. SIU 2022, 1-4.
Mehmet Haklıdır (2022). Deep Reinforcement Learning for Partially Observable Markov Decision Processes. Ph.D. Thesis, Control and Automation Engineering, Istanbul Technical University.
Mehmet Haklıdır, Hakan Temeltaş (2021). Guided Soft Actor Critic: A Guided Deep Reinforcement Learning Approach for Partially Observable Markov Decision Processes. IEEE Access, 9, 159672-159683.
DOI
Mahmut Lütfullah Özbilen, Emre Eğriboz, Ruşen Halepmollası, İsmail Bilgen, Mehmet Haklıdır (2021). Deep Reinforcement Learning for Simulation-Based Determination of COVID-19 Pandemic Mitigation Policies. Artificial Intelligence Theory and Applications, 1(2).
Mehmet Haklıdır, Isa Taşdelen (2009). Modeling, simulation and fuzzy control of an anthropomorphic robot arm by using Dymola. Journal of Intelligent Manufacturing, 20(2), 177-186.
DOI
Mehmet Haklıdır, Isa Taşdelen (2006). Modeling and Simulation of an Anthropomorphic Robot Arm by Using Dymola. IMS 2006.