Paleo-Inspired Robots Reveal How Extinct Species Moved

While decades of meticulous fossil study have revealed glimpses of Earth’s ancient inhabitants, the incomplete and fragmented nature of the fossil record continues to challenge our attempts to fully understand long-extinct species. Recently, researchers at the University of Cambridge have presented an innovative approach—paleo-inspired robotics—that seeks to emulate extinct animals in order to shed new light on their anatomy, movement, and evolutionary history.

What obstacles does traditional paleontology face, what breakthrough methods have the research team introduced, and how might these robotics-based reconstructions reshape our understanding of life’s complex journey through time?

The Complex Quest to Reconstruct the Past: The Challenges of Bringing Extinct Life to the Forefront

The field of paleontology has long fascinated scientists and the general public alike, with its discovery of fossils and remnants of ancient civilisations that have shaped the history of our planet. While significant advances have been made in the field, one major challenge continues to hinder our understanding of long-lost species – the incomplete and fragmented nature of the fossil record. 

Fossil evidence is often incomplete and partial, with many species known to us only from isolated remains. This lack of comprehensive data creates a daunting challenge for scientists attempting to reconstruct the anatomy of extinct animals. Gaps in the fossil record lead not only to uncertainty regarding the true nature of these organisms but also to a considerable amount of speculation regarding their appearance and behaviour. The absence of soft tissues and pigmentation poses a significant obstacle, making it difficult for scientists to accurately envision the creatures that once roamed the Earth.

Deciphering Movement: The Challenge of Ancient Locomotion 

In addition to the difficulties associated with reconstructing the appearance and behaviour of extinct species, understanding how these animals moved is also a major challenge. Fossils can offer clues regarding joint articulations and muscle attachment sites, but these findings are often insufficient to accurately recreate the biomechanics of extinct animals. The complexity of locomotion in ancient species can only be approximated through a series of assumptions, which may not always be accurate. As a result, scientists are frequently faced with incomplete or inaccurate reconstructions of the movements of long-lost species.

Another significant challenge in the field of palaeontology is the incomplete evolutionary pathways that led to the development of certain adaptations. The absence of transitional fossils can obscure our understanding of how specific traits emerged over time, making it challenging to trace the evolutionary history of a particular species. This lack of information can hinder our comprehension of the evolutionary processes that have shaped life on Earth and may even lead to incorrect assumptions regarding the origins of certain traits.

Researchers Explore the Possibility of Resurrecting Extinct Animals as Robots

A team of researchers from the University of Cambridge has published a review article in the journal Science Robotics outlining their vision for a new field of paleo-inspired robotics. The concept involves creating robotic systems that mimic the anatomy and movement of extinct animals with the goal of shedding light on evolutionary processes.

The concept of paleo-inspired robotics extends beyond reconstructing extinct species—it serves as a testbed for understanding the mechanical constraints that shaped their evolutionary trajectories. Unlike fossil-based reconstructions, robotic models can be adjusted iteratively, allowing scientists to assess how variations in limb structure, muscle placement, and articulation might have influenced locomotion and survival in different environments.

Using Robotics to Simulate Evolutionary Constraints 

By integrating insights from computational fluid dynamics and biomechanical modelling, researchers can explore the interaction between extinct creatures and their ecosystems in a controlled experimental setting. For example, evaluating the hydrodynamic efficiency of extinct marine reptiles such as plesiosaurs through robotic simulation can help determine whether their unique morphology provided an evolutionary advantage.

Understanding the intersection of natural and artificial evolution is crucial in paleo-inspired robotics. By examining how biological and robotic models interact, researchers gain deeper insights into the fundamental principles of movement, adaptation, and evolutionary transitions.

 Fig. 1. Life–artificial life loop. The study of natural life—through paleontology and biology—and artificial life—via bioinspired and paleoinspired robotics—offers complementary insights into animal evolution. The interaction between biology and robotics, as well as paleontology and paleo-inspired robotics, creates a two-way exchange of knowledge on the viability of different morphologies and movements. Just as paleontology and biology explore life’s natural progression, bioinspired and paleoinspired robotics examine how artificial evolution can parallel and inform these processes over time.

According to Dr. Michael Ishida, a co-author of the article and member of the University of Cambridge, paleo-inspired robotics can help researchers explore how changes to an animal's anatomy affect its movement, speed, and energy usage. By recreating extinct animals using robotics, scientists can simulate millions of years of evolution within a single day of engineering efforts.

The Role of Robotics in Experimental Paleontology 

This rapid prototyping of evolutionary change has profound implications for experimental paleontology. Traditional studies rely on comparative anatomy or fossil records, which can only provide static snapshots of evolutionary progress. However, by developing robots that mimic ancient locomotion patterns, scientists can analyse how small changes in morphology—such as the length of a tail or the positioning of limb joints—affect an organism’s overall biomechanical efficiency. This approach enables the experimental validation of long-standing hypotheses about evolutionary adaptations.

Furthermore, by drawing comparisons between paleo-inspired robotic models and their modern biological counterparts, researchers can bridge knowledge gaps in evolutionary biology. This method not only refines our understanding of extinct species but also informs bioinspired engineering, where evolutionary solutions from nature inspire modern robotic design.

Studying the Water-to-Land Transition with Robotics 

One example of a species that researchers are currently studying is the mudskipper fish. These fish have evolved the ability to "walk" on land, and the team hopes to build a robotic version of these fish to understand how they developed this ability. By analysing the mechanics and evolutionary pressures that led to the development of these fish, researchers can gain insights into how vertebrates shifted from aquatic environments to terrestrial habitats.

The study of the water-to-land transition is a prime example of how robotics can complement paleontological research. Previous studies have suggested that early vertebrates underwent a gradual adaptation to terrestrial environments, with intermediate forms displaying both aquatic and terrestrial traits. Using robotic simulations, researchers can systematically modify limb orientation, joint flexibility, and weight distribution to test how these factors influenced locomotion efficiency in different environmental conditions.

For instance, studies on modern amphibious fish, such as Polypterus senegalus, have shown that prolonged exposure to terrestrial habitats can induce skeletal and muscular changes that enhance land-based movement. By applying similar principles to robotic models of extinct species, scientists can explore whether certain evolutionary adaptations were driven by environmental necessity or opportunistic functional shifts.

Reconstructing Entire Extinct Creatures for Realistic Testing 

The team also hopes to recreate entire bodies of extinct animals, not just individual limbs or features. According to Dr. Ishida, analysing just one leg is not enough to understand how a four-legged creature walked, and robots can be placed in realistic environments to study movement and behavior. This approach has an advantage over computer simulations, which require complex models that capture physical properties such as sand or sticky surfaces.

Unlike purely digital simulations, which rely on complex mathematical models, robotic reconstructions provide tangible data on the real-world mechanics of extinct creatures. For example, the OroBot—a robotic model of the extinct stem amniote Orobates pabsti—was used to verify hypotheses about sprawling locomotion by replicating fossilised trackways. Such experiments enable researchers to test biomechanical feasibility in ways that traditional paleontological methods cannot.

Additionally, advancements in soft robotics and biomimetic materials allow researchers to simulate the flexibility of cartilage, ligaments, and soft tissues, elements that rarely fossilise but play a crucial role in locomotion. This holistic approach helps reconstruct not only the skeletal structure of extinct animals but also their potential range of motion and energy efficiency.

Unlocking Major Evolutionary Transitions with Robotics 

The researchers also believe that paleo-inspired robotics can shed light on large evolutionary transitions, such as the development of flight and the shift from quadrupedal to bipedal locomotion. By studying the anatomy of extinct animals and their movement patterns, scientists can gain insights into how these transitions occurred and what drove them.

One of the most compelling applications of paleo-inspired robotics is its ability to test theories of evolutionary convergence. Certain locomotor patterns—such as bipedalism—have evolved independently in multiple lineages, from theropod dinosaurs to early hominids. By constructing robotic models with varying centre-of-mass distributions, limb proportions, and gait mechanics, researchers can analyse whether these transitions followed similar biomechanical pathways.

Moreover, integrating machine learning algorithms with robotic models allows for real-time adaptation, simulating how evolutionary pressures might have guided specific anatomical changes. By training these robotic systems to optimise locomotion strategies under different environmental constraints, scientists can gain valuable insights into the selective forces that shaped the development of flight, quadrupedalism, and bipedalism.

Prof. Steve Brusatte, an expert in palaeontology from the University of Edinburgh, agrees with the researchers' vision for paleo-inspired robotics. He believes that these robots have the potential to revolutionise the field of palaeontology and help us understand the history of life on Earth. By recreating extinct species and studying their behaviour, researchers can gain a deeper understanding of how life evolved over millions of years.

Transforming Engineering Practices with a New Perspective

The study of fossilised anatomy is not only beneficial for palaeontologists but also has the potential to revolutionise engineering practices across various industries. By drawing inspiration from the mechanical design of ancient creatures, engineers may discover novel approaches to motion, balance, and efficiency that can be applied to modern technologies. The insights gained from studying the anatomy of extinct life forms can help engineers design more innovative and effective solutions, leading to significant advancements in fields such as robotics, biomechanics, and materials science.

By combining the strengths of different fields and perspectives, researchers can unlock new discoveries and insights that might not have been possible within a single discipline. The collaboration between paleontologists, engineers, and biologists can lead to a deeper understanding of the evolutionary processes that shaped life on Earth, as well as the development of innovative technologies that can benefit society as a whole.

As the technology behind paleo-influenced robots continues to evolve, it may push the limits of what we consider possible in the field of robotics. The ability to recreate extinct life forms and study their behaviour in controlled environments can lead to new insights into motion and adaptability, which can be applied to a wide range of applications. The future of robotics may hold more surprises than we can currently imagine, and the lessons learned from studying the past can help shape the direction of this field. By harnessing the power of paleo-influence, we may be able to unlock new possibilities for robotics and create a future that is both exciting and unpredictable.