MARVEL (short for Magnetically Adhesive Robot for Versatile and Expeditious Locomotion) can move quickly on metal floors, walls, and ceilings.

Image source: Interesting Engineering

Because of the potential dangers it poses to humans, MARVEL is designed for use in extreme circumstances. The research found that "Legged robots could potentially conduct various movements in constrained and complex 3D environments."

Purpose

It is a complex process to investigate industrial sites. For example, steel bridge construction and shipyard welding platforms are hazards to human employees due to their height and the close quarters in which they must operate. Researchers at Korea's Advanced Institute of Science and Technology have created a quadrupedal climbing robot equipped with magnetised feet to address this issue. Recent issues of Science Robotics feature their findings.

Capability

The team's biggest hurdle was figuring out the adhesion mechanism, which allows the robot to grip metal surfaces and facilitates movement in opposition to gravity.

Electromagnets and magnetic elastomers were used as the answer, enabling the robot to modify the magnetic characteristics of each of its feet.

MARVEL can also scan surfaces for openings and gaps. By modelling how a cat uses its front paws to scan items before going forward, researchers could design the robot to navigate around or over such obstructions.

According to the study's abstract, MARVEL completed a variety of testing situations, including "walking over a 10-centimetre-wide gap, overcoming a 5-centimetre-high obstacle."

The team has successfully tested the robot with a maximum payload of 6.6 pounds and speeds of up to 0.7 metres per second (m/s) (three kilograms).

Conclusion

The quadruped robot is intended to take the place of physical inspection by humans in dangerous environments or constrained places. For example, marvel can "examine industrial areas, such as steel-structured buildings, bridges, ships, or storage tanks," according to the report.

Based on a variety of carefully thought-out learning scenarios for Marvel to explore, the team has been able to achieve these achievements. Nevertheless, the researchers stated that "MARVEL will be able to climb and perform multiple jobs independently using a vision sensor, motion planner, and trajectory optimization."