معرفی
Dejan Milojevic is a Lecturer at the Department of Mechanical and Process Engineering at ETH Zürich, where he works in Prof. Emilio Frazzoli's group at the Institute for Dynamic Systems and Control. He received his Ph.D. in robotics from ETH Zürich with affiliations at both the Institute for Dynamic Systems and Control and the Chemical Energy Carriers and Vehicle Systems Laboratory at Empa. He also earned his BSc. and MSc. in Mechanical Engineering from ETH Zürich.
His educational background includes:
- Ph.D. in Robotics, ETH Zürich (affiliated with Institute for Dynamic Systems and Control and Empa)
- MSc. in Mechanical Engineering, ETH Zürich
- BSc. in Mechanical Engineering, ETH Zürich
Dr. Milojevic's research focuses on the co-design of embodied intelligence systems, with particular interest in sensor selection and placement for mobile robots. His work explores the interplay between perception, decision-making, and control in autonomous systems. He is also keen on advancing uncertainty estimation and explainable AI within machine learning frameworks. His research has practical applications in autonomous vehicles and robotics.
His recent publications demonstrate a consistent focus on optimizing the integration of perception systems and decision-making for mobile robots. The research shows how resource prioritization influences sensor choice, with cameras preferred for cost-effective designs while lidar sensors are chosen for better energy and computational efficiency. His work provides actionable information for designers of autonomous systems.
His notable scientific contributions include:
- CODEI framework for Co-design of Embodied Intelligence
- Research on occupancy queries to quantify perception requirements for motion planners
- Integer Linear Programming approach for sensor and algorithm selection
Dr. Milojevic has conducted research at Stanford University under Prof. Marco Pavone and has professional experience as a software engineer at Vay in Berlin, Germany. He has led significant research projects including the Automated Driving Sensor Test Vehicle project with multiple industry partners.
