The PedestriANS robot is capable of adjusting its walking style based on its surrounding environment, similar to a human when they’re moving around. (Image Credit: Ahmad et al)
Humans generally have more adaptive walking styles than robots, which is based on their ability to move in certain environments. For example, humans will slow down when the floor is slippery, change directions to avoid puddles, speed up, walk on uneven terrain, etc. Researchers at Osaka University in Japan have developed PedestriANS, a bipedal robot that’s capable of adjusting its walking style based on changes in the surrounding environment. This was achieved by using a system that connected the movements and configurations of both legs. The team published their findings in SAGE’s journal Adaptive Behaviors.
This new robot, PedestriANS, relies on an actuator network system (ANS) to function. The team used ANS in in the past, on projects which included robotic spines, arms and multi-legged robots. A single motor is responsible for its movements with its legs linked together through an ANS. The robot’s structure is also fairly simple, allowing it to easily adapt to changes in its environment. This is achieved by exposing physical characteristics of its legs and interactions between its components, which produces different walking behaviors.
The researchers plan on adding an AI system to the robot so that it will be able to autonomously adjust its walking behaviors without human intervention. (Image Credit: Ahmad et al)
The team assessed the performance of the robot through a number of experiments. They tested how shifts in its morphology affected its walking behavior by examining how it adapts to different ground materials, with a strong focus on walking style changes, speed, stability and its direction of movement.
After evaluating the data from the first experiment, the team improved the robot’s design so that it would be able to move around more efficiently with different connection patterns. The improvements allowed the team to evaluate how the robot adjusted its morphology while it was moving. In their second experiment, they found that PedestriANS was able to effectively adapt its walking style based on its surrounding environment.
This robot could be used in the future to carry out tasks that require it to be more mobile where the environment is uncontrolled and constantly changing. Even though the robot is capable of altering its walking behavior that’s suitable for its environment, it’s still not able to autonomously identify morphological adjustments that are more fitting to certain situations. To overcome this obstacle, the team will continue working on the robot by installing an AI system so that it will be able to function without the need for human input and guidance.
Have a story tip? Message me at: cabe(at)element14(dot)com
Top Comments