How Robots Know Where They Are

Have you ever told someone to “meet me at the coffee shop on 3rd Street”?

You gave them coordinates: a street name and a building. Robots need the same thing. But they use numbers instead of street names.

This guide explains how robots know where they are — and where everything else is.

The Big Idea: Every Robot Needs a Map

Imagine you’re in a new city. You open a map app on your phone. The app shows:

Where YOU are (the blue dot)
Where the restaurant is (a pin)
The path between them

Robots do the exact same thing. But their map is made of numbers, not pictures.

Three Types of Maps Robots Use

Map 1: The World Map (Fixed in Space)

This is the “big map.” It never moves.

Think of it like the GPS coordinates of a factory floor. The corner of the building is always at the same spot. The robot’s charging station is always at the same spot.

Example:

The robot’s base is at (0, 0, 0)
The table is at (2, 3, 0)
The shelf is at (5, 1, 2)

These numbers never change. Even if the robot moves, the world map stays the same.

Map 2: The Robot’s Body Map (Moves With the Robot)

This map is attached to the robot. Wherever the robot goes, this map goes too.

Think of it like a backpack with a GPS. Even if you walk around, your backpack always knows it’s “on your back.”

Why this matters: When the robot’s arm moves, the hand’s position is measured from the robot’s body — not from the factory floor.

Map 3: The Tool Map (Attached to the Hand)

This map is attached to whatever the robot is holding.

Think of it like holding a phone. The phone knows where its screen is, relative to your hand.

Why this matters: If the robot grabs a pen, it needs to know where the pen tip is — not just where the hand is.

How It Works: A Simple Example

Imagine a robot arm with two joints:

        [Hand]  ← the gripper
          |
      Joint 2  ← elbow (bends)
          |
      Joint 1  ← shoulder (spins)
          |
        [Base]  ← sits on the floor

The robot needs to know:

Where is the base? (World map)
Where is Joint 1, relative to the base? (Body map)
Where is Joint 2, relative to Joint 1? (Body map)
Where is the hand, relative to Joint 2? (Tool map)

To find the hand’s position in the world:

Hand in world = Base position
                + Joint 1 position
                + Joint 2 position
                + Hand position

This is called adding up the chain. Every robot does this, whether it has 2 joints or 20.

Why This Is Hard (And Why It Matters)

Problem 1: Joints Move

When Joint 1 spins, everything above it moves too. Joint 2, the hand, the tool — all of them.

The robot must recalculate everything, every time a joint moves.

A factory robot might move 100 times per second. That’s 100 recalculations per second.

Problem 2: Maps Must Match

The robot sees the world through cameras and sensors. But those sensors use the world map. The robot’s arm uses the body map.

The robot must constantly convert between maps.

Example:

Camera sees a cup at world position (3, 2, 1)
Robot calculates: to grab it, the hand must be at world position (3, 2, 1)
Robot figures out: to put the hand there, Joint 1 must be at 45 degrees and Joint 2 at 30 degrees
Robot moves the joints

This conversion happens in milliseconds.

Real Robots and Their Maps

Robot	How Many Maps?	Why So Many?
Factory welding arm	3-4	Base, arm joints, welding tip
Self-driving car	10+	GPS, cameras, lidar, radar — each has its own map
Surgical robot	5-7	Patient body, instruments, camera, doctor’s controls
Drone	6+	GPS, body, camera, gimbal, target, home position

More maps = more complexity = more things that can go wrong.

Key Terms (Simple Definitions)

Term	What It Means
World frame	The “big map” — fixed in space, never moves
Body frame	The map attached to the robot’s body
Tool frame	The map attached to the robot’s hand or tool
Joint frame	The map attached to one joint
Transform	Converting a position from one map to another
Origin	The “zero point” of a map — like (0,0) on a graph
Axes	The X, Y, Z directions — like up/down, left/right, forward/back

Try It Yourself

You don’t need a robot to understand this.

Exercise 1: Your Body’s Maps

Stand still. Your feet are at your “world origin.”
Raise your right hand. Where is it, relative to your feet?
Bend your elbow. Where is your hand now?
You just did forward kinematics in your head.

Exercise 2: Paper Robot

Draw a robot arm on paper: base, two joints, hand
Label each joint with an angle (like “Joint 1: 45°”)
Measure where the hand is on the paper
Change Joint 1 to 90°. Where is the hand now?
You just calculated forward kinematics with a ruler.

What’s Next?

Now you know how robots track positions. The next step is learning how they calculate those positions — the math that happens inside the robot’s computer.

That math is called kinematics. It’s just geometry with a fancy name.

This is part of Robot Basics — our simple guides to how robots work.

How Robots Know Where They Are: A Simple Guide to Robot Coordinates