Hey friend, I still remember the time I opened a humanoid robot specification sheet and saw the number “42 degrees of freedom.” My immediate reaction was: “Why on earth does it need many joints? Wouldn’t 10 or 15 be enough?”
After years of watching robots like Tesla Optimus, Figure 01 Boston Dynamics Atlas and Unitree G1 I finally understand why. Today I want to explain Degrees of Freedom in a way no complicated math, just the real reasons humanoid robots need 20 to 40+ joints.
What Does “Degrees of Freedom” Actually Mean?
A degree of freedom is one way a robot can move, think of it like this: A simple door hinge has 1 degree of freedom it can only. Close, your shoulder has 3 degrees of freedom you can swing your arm forward and back left and right and rotate it, your wrist has another 3 degrees of freedom for orientation.
Each joint in a robot usually adds one degree of freedom.
Why Humanoids Need So Many Joints
A basic humanoid robot breaks down like this:
– Legs: 6 degrees of freedom per leg times 2 equals 12 degrees of freedom
(3 at hip 1 at knee 2 at ankle sometimes plus toes)
– Arms: 6–7 degrees of freedom per arm times 2 equals 12–14 degrees of freedom
– Torso: 2–3 degrees of freedom (waist rotation and bending)
– Neck and Head: 2–3 degrees of freedom (looking around naturally)
– Hands: 10–20+ degrees of freedom (highly dexterous fingers in advanced models)
That quickly adds up to 25–45 degrees of freedom in modern humanoids.
Tesla Optimus has around 40 degrees of freedom.
Figure 01 pushes higher dexterity in the hands and arms.
Boston Dynamics’ electric Atlas is designed with agility using many joints for dynamic movements.
Serial vs Parallel Chains
There are two ways to arrange joints:
– Serial chains (most common in humanoids)
Joints are connected one after another like links in a chain (shoulder to elbow to wrist).
The advantage is: workspace the hand can reach far and in many directions.
The disadvantage is: Errors add up along the chain and the arm becomes heavy and less stiff.
– Parallel chains (used in some parts)
Multiple links work together at the time (like some ankle designs or Delta robots).
The advantage is: Much stiffer and more precise.
The disadvantage is: workspace.
Most humanoids use chains for arms and legs because we need long reach and human-like flexibility. Some advanced designs mix both for example using mechanisms in the ankle for better stability while keeping serial chains in the arms.
Redundancy Having More Joints Than You “Need”
Here’s where it gets interesting.
To move a hand to a point in 3D space you only need 6 degrees of freedom (3 for position plus 3 for orientation).
Most humanoid arms have 7 or more degrees of freedom. That extra joint is called redundancy.
Why add joints? Because redundancy gives the robot superpowers:
– It can reach the point in many different ways (avoiding obstacles)
– It can keep the elbow in a position while reaching
– It can maintain balance while moving the arms
– It allows more natural human-like movements
Figure 01 often uses its redundant arms to reach around objects or adjust posture smoothly. Tesla Optimus does the same when working in factory spaces.
Workspace Analysis, The Reachable Area
Every robot has a workspace the volume of space its hand (or foot) can actually reach.
With 6 degrees of freedom the workspace can have holes or awkward gaps.
With 7+ degrees of freedom the workspace becomes much larger and more flexible. The robot can move its hand through paths without getting stuck.
This is especially important for legs. A leg, with 4 or 5 degrees of freedom would struggle to walk on uneven ground or climb stairs. Giving it 6+ degrees of freedom (plus toes) dramatically improves its ability to adapt.
My Personal Take
At first all those joints look like overkill. But once you understand the physics of balance, manipulation and working in human environments you realize they’re not a luxury they’re a necessity.
We already talked about how important toes, compliant joints and good balance’re. All of those things become much easier when the robot has degrees of freedom to adjust its whole body smoothly.
Nature gave us 244 degrees of freedom in the human body (counting every small joint). Robots are still behind but 30–45 degrees of freedom is the current sweet spot where they become truly useful without becoming impossible to control.
The more I watch Tesla Optimus, Atlas and Figure improve the more I appreciate this balance: enough joints to be capable but not so many that the robot becomes too heavy, expensive or hard to control.