Papers
35
Total Citations
481
H-Index
13
About
Toshiyuki Kondo is a robotics and artificial intelligence researcher whose work has made significant contributions to autonomous mobile robot control, bio-inspired computing, and neural network-based learning systems. His research is most prominently centered on the application of artificial immune networks to robot behavior arbitration — a novel approach addressing one of behavior-based AI's longstanding challenges: how to dynamically and adaptively coordinate competing behavioral modules in unpredictable environments. Beginning in the mid-1990s, Kondo pioneered the "Immunoid" architecture, a decentralized framework inspired by biological immune systems that enables robots to flexibly arbitrate behaviors without rigid pre-programmed hierarchies. This foundational work, developed across a prolific cluster of publications around 2002, has collectively garnered over 150 citations. He further extended his contributions by integrating reinforcement learning with evolutionary state recruitment strategies (69 citations) and exploring neuromodulatory neural networks for robot control (35 citations). His work on seamless simulation-to-real-world transfer using dynamically rearranging neural networks (36 citations) reflects a practical engineering sensibility alongside theoretical innovation. More recently, Kondo has explored visuomotor learning through passive motor experience, with implications for neurorehabilitation and brain-computer interface research, demonstrating a career-long commitment to bridging biological inspiration with real-world intelligent systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8Visuomotor learning by passive motor experience26 citations · 2015
- 9
- 10