MagicLab, an innovative startup specializing in humanoid robots, has unveiled an interesting video showcasing their latest creation, MagicBot.
In this video update, the humanoid impressively roasts marshmallows executes magic tricks and even busts a move on the dance floor. The footage represents the robot's remarkable dexterity and adeptness in manipulating objects.
In January, MagicLab revealed its humanoid design, positioning it as a formidable competitor to Boston Dynamics' agile Atlas robot. The unveiling showcased its exceptional balance and dexterity, highlighted by its impressive capability to execute somersaults.
Impressive skills
In the video's opening scene shared by MagicLab, the robot, controlled remotely, grabs a skewer to pierce a marshmallow, then carefully roasts it over a barbecue grill. Next, MagicBot neatly folds baby clothes and places them in a basket.
The robot surprises everyone by dancing well to upbeat music and watering a plant with delicate touches on its leaves, showcasing human-like care and precision. Delving into its magician role, MagicBot transforms an egg into a ball, showcasing its impressive skill in handling small objects and its adaptability to various everyday tasks with smoothness and agility.
The skills were made possible using a blend of miniature high-torque servo actuators and sensitive multi-dimensional pressure sensors that helped the team craft an exceptionally dexterous hand.
According to the developers, the six miniaturized servo actuators are instrumental in enabling a wide array of object manipulations, facilitating users' intuitive and swift handling experience.
With sub-millimeter positioning accuracy and a load capacity reaching several kilograms, this hand can adeptly grasp objects of diverse shapes, sizes, weights, and substances, including delicate, slippery, or malleable items.
The team asserts that the hand can mimic approximately 70 percent of all human hand gestures, hinting at its vast potential across various domains.
Its versatility promises to revolutionize numerous fields, from specialized industrial tasks to service applications, spanning hazardous material handling, public service, and domestic operations.
Advanced proposition
Conventional bipedal humanoid robots encounter significant hurdles when attempting stunts like front flips or somersaults due to power limitations and balancing issues.
MagicLab, a recent entrant in the field, devised custom motor joints and actuators to surmount these challenges, allowing aerial maneuvers previously exclusive to hydraulic systems like Atlas.
MagicLab's humanoid model integrates proprietary D190 joint motor modules, delivering impressive torque capabilities. These modules, paired with specialized actuators, enable the robot to handle voltage fluctuations and achieve precise manipulation, according to the startup.
The strategic weight redistribution toward the thighs and hips enhances the robot's dynamic stability and maneuverability during leaps and jumps.
The conpany also asserts that its electrically driven design matches and surpasses hydraulic counterparts in performance while offering cost efficiency, compactness, and a lighter build.
Moreover, the electric drive system reduces energy consumption and minimizes environmental risks associated with fluid leaks. MagicLab claims this innovative approach heralds a new era in humanoid robotics, where agility, affordability, and sustainability converge seamlessly.
MagicLab envisions its humanoid model as a versatile asset for deployment across various scenarios. Its agility enables it to maneuver through complex terrains, making it suitable for search and rescue missions where navigating obstacles is crucial.
Beyond emergency response, the robot holds promise in diverse fields such as sports, entertainment, industrial automation, logistics, and transportation. By enhancing work efficiency and alleviating human workload, it aims to revolutionize these sectors with its multifaceted capabilities, claims the company.
Originally published on Interesting Engineering : Original article