Robotic Paleontology A new Key to Understanding Early Mammal Evolution

• Applicant: Professor Dr. John A. Nyakatura
• Subject Area: Palaeontology ; Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 560044335

Placental mammals have evolved increased resting metabolic rate, improved cognition, intensified and highly specialized parental care, complex communication, effective locomotion, and re-configured their lung ventilation as well as their dentition. In combination, these aspects (among others) set them apart from other vertebrates. While this suite of features is interdependent (e.g., increased cognitive capabilities necessitate a heightened metabolic rate, which in turn requires effective breaking down of ingested food) and is often difficult to pinpoint in the fossil record, the ‘sprawling-to-erect-paradigm’ and the fundamental re-configuration of the locomotor apparatus facilitating typical mammalian locomotor behavior became emblematic for this decisive, macroevolutionary transition in the forerunners of extant placentals. However, the classic linear narrative of a progression from sprawling limb posture towards erect posture and locomotion still found in most textbooks has been identified as deficient and inadequate to explain the observed diversity in the fossil record. Without teleological arguments, a linear transition has not been satisfyingly explained. Understanding the adaptive benefits that drove the evolution of stem and early crown mammals requires a mechanistic understanding of limb function in transitional forms within an up-to-date evolutionary framework. In previous interdisciplinary work, our group of collaborators helped to establish the field of ‘robotic paleontology’, a rigorous paleontology-simulation-robotics loop, which by integrating methods from diverse fields allows studying the biomechanics of key vertebrate fossils using quantitative hypothesis testing. While this method has proven able to successfully reconstruct locomotor mechanics, which can then be traced along a phylogenetic tree, it relies on an elaborate blend of expertise and methodological approaches that need to be addressed in custom-designed cross-disciplinary settings. I here propose to leverage my network of excellent researchers in the relevant fields to set up an exceptional robotic paleontology research environment in order to provide a new view on early mammalian evolution.

• DFG Programme: Reinhart Koselleck Projects