Rafael Kiesel

Papers

Total Citations

H-Index

About

Rafael Kiesel’s research lies at the intersection of evolutionary robotics, embodied intelligence, and adaptive control systems. His most influential work explores how robots with evolving morphologies can learn to control their bodies through lifetime adaptation, drawing inspiration from biological principles. In his highly cited 2017 paper (9 citations), Kiesel analyzed Lamarckian evolution in morphologically evolving robots, demonstrating that newborn robots must acquire new controllers to match their inherited body plans—a process that cannot rely solely on parental control strategies. He further advanced this concept in a subsequent study (2 citations), where he implemented online evolution as a form of lifetime learning. To achieve this, Kiesel developed an innovative indirect encoding scheme combining Compositional Pattern Producing Networks (CPPNs) with Central Pattern Generators (CPGs), enabling modular robots to generate effective open-loop gait controllers. This work provides a foundational framework for understanding how evolution and learning can interact in embodied systems, with implications for adaptive robotics and artificial life. Kiesel’s contributions are particularly notable for bridging the gap between evolutionary algorithms and real-world robotic adaptation, offering a pathway toward more resilient and self-improving machines.