Kohei Nakajima

About

Kohei Nakajima is a pioneering researcher at the intersection of soft robotics, embodied intelligence, and neuromorphic computing, whose work has fundamentally reshaped how scientists think about computation, control, and physical bodies in robotic systems. His most influential contribution lies in developing the concept of "physical reservoir computing," demonstrating that the dynamic properties of soft, compliant bodies — rather than being engineering liabilities — can serve as powerful computational resources. His landmark 2013 study on octopus-inspired soft robotic arms (176 citations) showed that control traditionally delegated to a central brain can be effectively "outsourced" to the body itself, a concept now central to morphological computation research. Nakajima's work spans biological inspiration broadly: from quadruped robots exploiting spine dynamics for locomotion, to flapping-wing drones that naturally detect wind direction through their own structural movement. His 2022 work on physics-informed recurrent neural networks for soft pneumatic actuators further bridges machine learning and physical modeling to advance indirect sensing in flexible robots. Across more than a decade of research, his cumulative citation record reflects growing community recognition that intelligence is not solely brain-bound — a foundational insight driving the next generation of adaptive, embodied robotic systems.

Research Focus

Key Achievements

Top Papers

1. 1
   A soft body as a reservoir: case studies in a dynamic model of octopus-inspired soft robotic arm

   176 citations · 2013
2. 2
   Physics-Informed Recurrent Neural Networks for Soft Pneumatic Actuators

   70 citations · 2022
3. 3
   Spine dynamics as a computational resource in spine-driven quadruped locomotion

   57 citations · 2013
4. 4
   Computing with a muscular-hydrostat system

   41 citations · 2013
5. 5
   Flapping‐Wing Dynamics as a Natural Detector of Wind Direction

   39 citations · 2020
6. 6
   Embodiment enables the spinal engine in quadruped robot locomotion

   35 citations · 2012
7. 7
   Timing-based control via echo state network for soft robotic arm

   29 citations · 2012
8. 8
   Octopus-inspired sensorimotor control of a multi-arm soft robot

   29 citations · 2012
9. 9
   Behavior switching using reservoir computing for a soft robotic arm

   29 citations · 2012
10. 10
    BOOTSTRAPPING PERCEPTION USING INFORMATION THEORY: CASE STUDIES IN A QUADRUPED ROBOT RUNNING ON DIFFERENT GROUNDS

    28 citations · 2012

Key Collaborators

Contact & Links