In April 2024, the company announced that they had retired the hydraulic based HD Atlas in favour of a new all electric 5th generation version of Atlas. Atlas is demonstrating policies developed using reinforcement learning with references from human motion capture and animation. This work was done as part of a research partnership between Boston Dynamics and the Robotics and AI Institute (RAI Institute).
Atlas, Boston Dynamics' most dynamic humanoid robot, is designed with a focus on advanced mobility and dexterity. This robot is fully electric and utilizes an intricate actuator system, making it capable of complex, agile movements. Atlas is equipped with 28 joints and advanced control algorithms that enable it to execute tasks with high precision and adapt to its environment. The robot’s design allows it to perform physically demanding tasks like running, jumping, and even executing backflips, demonstrating significant advancements in humanoid robotics.
Atlas is notable for its construction materials and structural design, which emphasize a high strength-to-weight ratio. The robot incorporates titanium and aluminum, extensively using 3D printing technology to optimize its frame for robustness while maintaining a relatively low weight of 89 kg. This engineering approach grants Atlas the ability to perform dynamic movements and actions, such as somersaults and balancing tasks, which require both strength and lightweight components to maintain high levels of performance and efficiency.
The potential applications of Atlas extend beyond demonstrations of agility and power. Boston Dynamics leverages Atlas to explore the practical uses of humanoid robots in real-world scenarios that require a blend of human-like mobility and robotic precision. While Atlas is primarily a platform for research and development, its capabilities hint at future roles in scenarios like disaster response, where navigating hazardous or complex terrains is necessary, or in jobs that require high levels of mobility and manipulation skills. Atlas represents a significant step toward the integration of humanoid robots into everyday tasks and environments.
Boston Dynamics - ATLAS 003 Specs
1 - POWER - Fully electric power and actuation
2 - Product mass - 89 Kg
3 - Top Speed - 2.5 mtrs\sec
4 - Height - 1.5 mtr
5 - Battery Capacity: 3.7 kWh lithium-ion
6 - Three onboard computers handle control, perception, and estimation
7 - 1,000 sparse 8 bit integer TOPS + 1 trillion sparse 8 bit integer TOPS
The Atlas robot is powered by a 3.7 kWh lithium-ion battery pack. This battery
can provide roughly one hour of "mixed mission" operation. The fully electric
Atlas 003 has non hydraulic all electric actuators
Atlas uses the Nvidia's Jetson AGX Thor computing platform with 2560 CUDA
cores and OS for advanced AI processing, and an early adopter of Nvidia's
Isaac GR00T platform
Features - Agility, AI, arms, hands, legs, manipulation, mobility, vision
articulated head with LED lights, and partially autonomous capabilities
MIT Leg Lab and Boston Dynamics History
The MIT Leg Laboratory was founded by Marc Raibert in 1980. He directed the lab until 1995. It was originally established when Raibert was the Associate Professor at CMU’s Computer Science and the Robotics Institute. When he became a Professor of Electrical Engineering and Computer Science at MIT in 1986, he then moved the Leg Lab there and named it the MIT Leg Lab.
The MIT Leg Laboratory explores active balance and dynamics in legged systems, robots and animals alike. simulating and building creatures which walk, run, and hop like their biological counterparts. Systems that balance actively can use footholds that are widely separated, they can move along narrow paths with a narrow base of support, and they can use their kinetic energy as a bridge to increase their effective size. Dynamics is a key ingredient in the behaviour of animals and it will weigh heavily in the development of useful legged vehicles.
The techniques used to control each of the running machines derive from a single simple set of control algorithms. They focus on support, posture, and propulsion. These algorithms have been adapted for hopping, pronking, biped running, fast running, trotting, pacing, bounding, and gymnastic maneuvers. The ability of simple algorithms to operate under these diverse circumstances suggests their fundamental nature.
The "MIT Leg Lab" Team
Your browser does not support the video tag.
Free Hit Counter supplied from www.hitwebcounter.com - CLICK HERE
You are guest number
Since 08\05\2025
Please visit again
IN-AP Systems - Established 1997
A computer is like the mind, it has IN-finite AP-plications. . .
