The field of robotics is advancing by leaps and bounds, and one of the areas of greatest development is locomotion. Legged robots and wheeled robots have traditionally dominated this field, each with their own advantages and drawbacks. However, a team of researchers at ETH Zurich’s Robotic Systems Lab has challenged this dichotomy by developing a new autonomous system that combines the versatility of legs with the efficiency of wheels.
This innovative robot, equipped with four legs and a wheel on each, represents a significant advance in the search for agile and robust robots capable of navigating complex terrain. The combination of both locomotion systems allows the robot to adapt to a variety of surfaces, from uneven terrain to smooth surfaces, with impressive efficiency and speed.
La clave de este avance reside en el desarrollo de una red neuronal artificial altamente sofisticada. Esta red neuronal, el «cerebro» del robot, es capaz de procesar simultáneamente información proveniente de múltiples sensores, incluyendo cámaras, sensores de movimiento y sensores táctiles. Esta información sensorial se traduce en una comprensión detallada del entorno del robot, permitiéndole identificar obstáculos, evaluar la superficie del terreno y planificar rutas óptimas en tiempo real.
The processing speed of the neural network is crucial to the success of the system. With travel speeds reaching 20 km/h, the robot needs to make navigation decisions in fractions of a second. A delay of just half a second at this speed could result in a disastrous collision.

To address this challenge, researchers have optimized the neural network architecture and developed advanced control algorithms that allow the robot to react with the necessary speed and precision. The system is capable of evaluating different movement options in milliseconds, selecting the safest and most efficient action for each situation.
Tests carried out with a prototype of the robot have shown promising results. These tests were carried out in the cities of Zurich and Seville with very positive results.
The robot has demonstrated its ability to navigate challenging terrain, avoiding obstacles and adapting its mode of locomotion to the characteristics of the terrain. Additionally, a significant improvement in energy efficiency has been observed compared to robots that use only legs or wheels.
This advance in robotic locomotion opens a range of possibilities in different fields. Robots with these capabilities could be used in search and rescue tasks in natural disasters, space exploration, infrastructure inspection and even in assisting people with reduced mobility.
Despite progress, there are still challenges to overcome. Optimizing energy consumption, developing more robust perception systems, and improving robot autonomy are just some of the challenges that researchers continue to address.
However, the development of this autonomous system for legged and wheeled robots represents a significant step towards creating more versatile, robust and efficient robots. In the not-too-distant future, these hybrid robots could become an important tool in a wide range of applications, transforming the way we interact with the world around us.
A video can be seen at: https://youtu.be/vJXQG2_85V0