The Challenges in Building Humanoid Joints Part 2: Arms & Neck
In this continuation of our series exploring the intricate world of humanoid robot joints, we shift our focus from the legs to the upper echelons of robotic design: the arms and neck. If the legs are facing the challenge of keeping balance. The arms face two kinds of challenges: 1 being able to swing 2 being able to handle a heavy payload. The neck of a humanoid robot presents a different set of challenges, primarily focused on the ability to express human-like emotions and facilitate human-robot interaction.
1. Arms – The challenge of swinging and carrying
Designing humanoid robot arm joints poses challenges on two distinct levels. Firstly, there’s the task of replicating the natural swing action observed in human arms. Secondly, there’s the necessity to ensure the arms can handle substantial payloads, enabling the robot to perform various tasks such as those encountered in warehouses or construction sites.
Several studies indicate that the swinging action of arms has also stabilizing effects on the human walk. Furthermore, other recent studies suggest that the swing of arms decreases the metabolic cost of human locomotion. Although it is a trivial task for humans to swing their arms, comprehensive methods are required for humanoid robots to utilize arms for walking stabilization and control purposes. This difficulty in replicating very easy tasks for humans is called the ”Moravec’s paradox”.
a. Shoulder:
For humanoid robots, the shoulder joint is another complex area that requires a significant range of motion to replicate human-like movements. In humans, the shoulder has three main degrees of freedom (DOF):
- Flexion/Extension Actuator: This allows the robot’s arm to move forward and backwards.
- Abduction/Adduction Actuator: Enables the robot’s arm to move sideways, away from and towards the body.
- Internal/External Rotation Actuator: Facilitates rotational movement of the arm around its axis.
- Transverse Abduction/Adduction Actuator: Allows for the movement of the arm across the front of the body and back out to the side ( not in all robots)
Achieving such movements typically requires multiple actuators. For example, a common setup includes:
- One actuator for flexion and extension,
- A second actuator for abduction and adduction,
- A third for internal and external rotation.
- A possibly fourth actuator for transverse movement.
Thus, the minimum number of actuators in a humanoid robot’s shoulder joint to achieve basic human-like movement is typically three. Still, the exact number can vary depending on the specific requirements and goals of the robot’s design. More advanced or specialized robots may incorporate additional actuators to achieve greater flexibility and range of motion.
b. Elbow:
Elbow joints utilize compact servo motors or hydraulic pistons. These components deliver controlled flexion and extension motions with minimal backlash, prioritizing robustness, and precision to effectively manage varying loads and tasks encountered in real-world environments.
It’s important to note that the weight of the actuator in the elbow can have a cascade effect on the entire arm system, making it heavier to move the wrist and potentially impacting overall arm performance. Therefore, lightweight yet strong materials and efficient actuator designs are crucial to minimize this cascade effect and ensure optimal performance across all joints.
c. Wrist:
Among all the humanoid joints in the arm, wrist joints face a unique challenge. They often need to handle substantial payloads while maintaining stability and accuracy. Paradoxically, to achieve greater precision and faster manipulation of heavier objects, wrist joints require stronger actuators. However, the inclusion of heavier actuators at the end of the arm can lead to a decrease in the arm’s overall load-carrying capacity.
Wrist joints typically employ compact rotary actuators or servomotors to enable precise articulation for manipulation tasks. These components are crucial for maintaining responsiveness and accuracy in hand movements. Lightweight construction is essential to ensure that the additional weight of the actuators does not compromise the arm’s maneuverability or increase energy consumption. Addressing the constraints of weight and space presents a key challenge in the design of wrist joints for humanoid robot arms. Balancing the need for strength and precision.
Output: Integrating the Archimedes Drive into actuators for both the shoulders, wrist, and elbow joints of a humanoid robot significantly boosts precision and power. With minimal backlash, movements are smoother and more accurate, crucial for the fine motor skills needed such as at the wrist. Meanwhile, the high torque density ensures robust and efficient elbow operations, enhancing the robot’s ability to perform tasks requiring both strength and delicate manipulations.
2. Neck – the challenge of mimicking human expression
The neck joint is gaining importance as humanoid robots strive to mimic human expressions, a necessary skill to integrate robots in human working space. The ability of a robot to tilt its head in curiosity, nod in understanding, or swivel to maintain eye contact transforms it from a mere machine into a relatable entity. Neck joints integrate lightweight rotary actuators or servo motors to enable rotation and tilting of the head. Design considerations include compactness and precision to facilitate fluid head movements while minimizing overall weight.
Furthermore, as robots are interacting more and more with various environments, the neck actuator must be composed of sensors – its cameras, microphones, and other input devices. We need to ensure that the movements of the neck actuators enhance the robot’s sensory input organs without causing disorientation or misalignment of the sensors.
However, emulating the human neck presents challenges due to its complexity. A human neck comprises nine vertebrae, each providing six degrees of freedom (DOF), making replication particularly intricate.
a. Types of Neck Actuators
Electric Motors: Compact and precise, electric motors are commonly used for neck actuation. They can be servo motors, which offer fine control over movement and position, making them ideal for nuanced expressions and accurate head orientation.
Pneumatic Actuators: Less common in neck applications, pneumatic actuators use compressed air to create movement. While they can provide smooth motion, their use is often limited by the need for air supply systems and less precise control compared to electric motors.
Hydraulic Actuators: Known for their ability to exert strong forces, hydraulic actuators are less frequently used in humanoid necks due to their bulkiness and the complexity of hydraulic systems. Furthermore, hydraulic systems can leak, posing a risk in environments that are sensitive to contamination or where electronics are present. However, they might be considered in larger robots requiring significant head support.
Shape Memory Alloys (SMAs): SMAs offer unique advantages in creating more human-like movements due to their ability to contract and relax in response to electrical currents, mimicking muscle action. Their application in neck actuators is still emerging, with the potential for highly natural movements.
b. Key Considerations for Neck Actuators
Range of Motion: To mimic human head movements, the actuators must allow for a multi-axis range of motion, including pitch (nodding), yaw (shaking), and roll (tilting).
Load Capacity: The actuator must support the weight of the robot’s head, which can vary significantly depending on the sensors and components housed within it.
Speed and Precision: Quick, precise movements are necessary for responsive and expressive interactions, requiring actuators that can adjust swiftly and accurately.
Integration with Sensors: The neck often plays a crucial role in positioning sensors (e.g., cameras, microphones). Actuators must work seamlessly with these components to orient them appropriately for tasks like face recognition, environmental scanning, and interaction with humans.
Durability and Safety: Given their essential role in interaction and communication, neck actuators must be reliable and designed with safety in mind, ensuring smooth operation without sudden or jerky movements.
Various solutions have emerged to address this challenge. Some manufacturers, for instance, have opted to incorporate additional actuators into the neck, to allow robots to convey emotions like confusion. Nonetheless, they remain a considerable challenge for robot developers. Balancing the need for expressive movement with mechanical stability requires innovative engineering solutions and ongoing refinement in humanoid robotics.
