Pasi Koikkalainen

Lappeenranta-Lahti University of Technology, University of Jyväskylä

Papers

Total Citations

H-Index

About

Pasi Koikkalainen is a pioneer in the application of self-organizing neural networks to robotics and autonomous systems. His research centers on how machines can learn motion behaviors and generate intelligent action plans directly from sensory data, without explicit programming. Koikkalainen’s most influential work, “Self-Organization and Autonomous Robots” (1997, 17 citations), laid foundational ideas for enabling robots to adapt their behavior through unsupervised learning. He is perhaps best known for developing a tree-structured variant of the self-organizing map (SOM) that elegantly solves a critical industrial challenge: automated surface processing. His 1996 paper on this method (6 citations) demonstrates how a robot can learn to generate optimal tool paths for painting, coating, or sandblasting from an arbitrary set of surface data points. This hierarchical, deterministic SOM approach not only simplified path generation but also improved precision and adaptability in manufacturing. Koikkalainen’s work bridges theoretical neuroscience-inspired learning with practical robotics, showing how self-organization can transform raw sensor data into purposeful, efficient action. His contributions remain relevant for researchers in autonomous robotics, neural computation, and intelligent manufacturing.