Vince Jankovics

Papers

Total Citations

H-Index

About

Vince Jankovics is a robotics and artificial intelligence researcher whose work bridges the gap between physical manipulation and intelligent decision-making. His research focuses on three key areas: compliant robot control, tactile sensing, and reinforcement learning. Jankovics’s early contributions include developing an artificial neural network-based compliant control system for robot arms (2016, 3 citations), which enables safer and more adaptive human-robot interaction. He also advanced tactile sensing technology by designing high-resolution sensors for curved robotic fingertips (2014, 2 citations), addressing a critical challenge in dexterous manipulation. More recently, Jankovics has ventured into deep reinforcement learning, proposing an efficient entity-based approach (2022, 2 citations) that moves beyond fixed-size inputs to handle complex, variable-length observational data. This work tackles a fundamental limitation of traditional DRL methods, opening new possibilities for robots operating in unstructured environments. While his citation counts are modest, Jankovics’s research trajectory demonstrates a thoughtful progression from hardware-level sensing to high-level learning algorithms, reflecting a comprehensive understanding of the robotics pipeline. His work is particularly relevant for researchers interested in integrating tactile feedback with learning-based control.