Agility Robotics' Digit Robot Demonstrates Advanced Control for Heavy Object Handling

Tangent, Oregon – Agility Robotics' humanoid robot, Digit, is showcasing advanced capabilities in manipulating heavy objects, underscoring the potential for bipedal robots in automating physically demanding tasks. Chris Paxton, AI Innovation Lead at Agility Robotics, recently highlighted Digit's robust control strategy, confirming the robot's ability to handle substantial loads, including cinderblocks.

"Yes those are cinderblocks! One of the advantages of this control strategy is that it's pretty simple to adapt to heavy objects, like we'd want to do for robots to be useful in the kinds of difficult jobs we want to automate," Paxton stated in a recent social media post. This emphasizes the company's focus on developing robots for challenging industrial environments.

Digit, which stands 5 feet 9 inches tall and weighs approximately 143 pounds (64.8 kg), is designed to operate in human-centric spaces and can lift objects weighing up to 35 pounds (16 kg). Its bipedal design allows it to navigate complex environments, including stairs and narrow aisles, where traditional wheeled robots might struggle. The robot's "backward" leg design contributes to its agility and ability to maintain balance.

A key to Digit's performance is its "whole-body control foundation model," a relatively small LSTM neural network trained extensively in simulation. This control system, likened to the motor cortex in the human brain, enables Digit to maintain stability and execute precise movements even when handling heavy payloads or encountering disturbances. The model is capable of zero-shot transfer from simulation to the real world, allowing for rapid deployment and adaptation.

Agility Robotics envisions Digit filling critical labor gaps in logistics and manufacturing, particularly for repetitive and strenuous tasks like tote handling. The company is preparing for large-scale production, with a goal of manufacturing up to 10,000 Digit robots annually. Initial commercial deployments, including pilot programs with Amazon and GXO Logistics, focus on tasks such as moving empty totes and transferring products between autonomous mobile robots and conveyor systems.

The development of Digit's advanced control strategies for heavy object manipulation marks a significant step toward the broader adoption of humanoid robots in industries requiring versatile and robust automation solutions. This capability is crucial for expanding the robot's utility in "difficult jobs" and fostering a collaborative environment where robots work alongside human colleagues.