Snakes: teacher of robotics

Following the example of snakes, researchers have developed a robot that can climb steps. (Image: JHU / Will Kirk)

They effortlessly meander through rugged terrain and overcome obstacles and steps – researchers have now gained new insights into the fascinating movement technology of the snakes. They were then able to successfully transfer these findings to a robot. Your results could now be used to develop meandering search and rescue robots that are becoming more and more similar to their natural role models.

The sight of a writhing snake shakes some people’s spines – but it is fascinating for scientists and technicians. The reptiles have developed sophisticated locomotion techniques that enable them to move in very different environments: they wind rapidly over sand dunes, meander through scree landscapes, overcome obstacles or climb trees. This enabled the snakes to become bizarre muses of robotics: Scientists are trying to transfer the flexible reptiles’ movement techniques to robots that are literally flexible. For example, a perfect robo snake could explore the debris during disaster operations.

On the trail of snake technology

So far, snake robots have been nimble on sandy ground and researchers have already “taught” their constructions the meandering climbing technique for climbing trees. But as Chen Li and Qiyuan Fu from Johns Hopkins University in Baltimore emphasize, a different ability of the natural role models is actually more important for search and rescue robots: they must be able to overcome step-like obstacles such as stone blocks. How the reptiles manage to do this has so far not been known in detail. As part of their study, the scientists first examined the snake technology when climbing stairs.

King snakes served them as visual objects. These snakes are masters of locomotion in complex terrain: “They regularly cross boulders and fallen trees in their habitat,” says Li. The research team carried out a series of experiments in which the snakes crossed steps of different heights and with different surface properties. The researchers used video recordings to capture in detail how the animals used their bodies in response to the challenges.

The observations showed that snakes divide their bodies into three sections when crossing steps. A front and a rear section remain flexible and move laterally when in contact with the horizontal areas. The middle section of the body stiffens to allow movement in the vertical to bridge the step. The moving areas ensure that the snake does not tip over during the action. In the process, the three functional segments move through the snake’s body, so to speak, the researchers say. The higher and slippery the steps were, the slower the test animals moved and their front and rear body sections meandered less to promote stability, the analyzes showed.

Robo snake climbing stairs

On the basis of this information, the researchers then designed a robot snake that mimics this movement pattern: it meanders forward on small wheels and can stiffen certain segments if necessary. Initially, the construction struggled to remain stable at higher levels – it wobbled, twisted, or got stuck on the levels, the scientists report. The technical adjustments then showed that a certain amount of flexible pressure was needed for the stability of the snakes: when the scientists equipped the robot snake with spring elements that sit between the wheels and the body, the stability increased significantly. In the end, the team was able to cross steps with a height that corresponded to up to 38 percent of their body length at snake-like speed, the team reports.

As the scientists emphasize, the energy supply for the snake robot is still a problem and the added spring elements further increased the demand. Nevertheless, they see their results as an important step forward: “The animal is still far superior, but these results have great potential in the development of snake robots that can overcome major obstacles,” says Li. He and his colleagues now want to continue tinker and develop models that can master even more complex 3D terrain with unstructured large obstacles.


Video: Patrick Ridgely, Dave Schmelick, Len Turner / Johns Hopkins University Office of Communications

Source: Johns Hopkins University, technical article: Royal Society Open Science, doi: 10.1098 / rsos. 191192

Recent Articles

Related Stories