Phase 01 · Week 1 · 105 minutes

Day 6: Coordinate frames, axes, origins, and handedness

Meet the robot, then meet the maths · Start with a physical robot and build every maths idea from counting, measuring, arrows, and grids.

Chapter 01 · Meet the robot, then build its mathematical language

Today in the field story

One problem, then the next

The camera is taped to the Parcel-Sorting Desk Robot while the parcel lanes are painted on the table. Both report valid numbers, yet their axes turn differently when the arm moves. Attach and name both maps, describe one parcel in each, and plant a reflected axis set. The exercise turns yesterday’s arrow into a statement whose frame can no longer be silently guessed.

Why now

Coordinates from different physical maps cannot be compared or added until their frames are explicit.

Ignore today

Ignore transform matrices and timestamp buffers; first make origins, axes, units, and handedness visible.

Unlocks next

Safe frame conversion, tf2 reasoning, and camera-to-robot calibration.

Understand

Build the physical picture first

A coordinate frame is a small map fastened to an object; when the object moves, its local map moves with it.

A coordinate frame has an origin and ordered axis directions. In 2D it usually has x and y; in 3D it has x, y, and z. Coordinates describe a physical point relative to that frame. A coin does not have one universal coordinate tuple: its table-frame coordinates and camera-frame coordinates may differ while the coin stays still. Always read the frame label before reading the numbers.

A frame differs from a point. The frame supplies the measuring map; the point is something described by that map. A robot base frame may be fixed to the chassis, a camera frame to the camera housing, and a world frame to the room or map. When the rover turns, camera axes turn with it while world axes remain fixed. This is why a local “one metre forward” vector changes direction in world coordinates.

Robotics normally uses right-handed 3D frames. Point the right-hand index finger along +x and the middle finger along +y; the thumb indicates +z for the usual mnemonic. More formally, x crossed with y points along z. A mirrored axis set reverses handedness and cannot be produced by an ordinary rigid rotation. You do not need to calculate a cross product today, but you must be able to detect a reflected drawing.

Never add or compare coordinate components merely because their arrays have the same shape. First ensure they describe compatible quantities in the same frame, unit, and time. A camera observation captured before the robot moved may be stale even after a correct frame conversion. Frame names and timestamps are therefore part of production data contracts, not optional comments.

Words you need

Name each idea precisely

Coordinate frame

An origin plus ordered axis directions used to describe coordinates.

Physical example:

A base frame drawn at the centre of a rover chassis.

Local coordinates

Coordinates expressed in a frame attached to the nearby object or sensor.

Physical example:

A cup located 40 cm straight ahead of a camera.

Reference frame

The named frame whose axes and origin are used for a description.

Physical example:

A map frame used to state all warehouse robot positions.

Right-handed frame

An axis ordering in which +x crossed toward +y gives +z.

Physical example:

x forward, y left, z up is one common robot-body convention.

Frame mismatch

Using coordinates as though they share axes and origin when they do not.

Physical example:

Treating a camera-forward offset as a map-x offset after the robot has turned.

Math, one line at a time

Work through today’s relationship

Prerequisite rescue · optionalStart at zero: counting, units, and coordinates

No maths is assumed. First give every robot number a physical meaning, a unit, a zero point, and a positive direction.

N
a count, such as number of jointsUnit: no unit
x
one position measured from a chosen zeroUnit: centimetres (cm) or metres (m)
Δx
change in x; Δ means changeUnit: same unit as x

  1. Draw a number line. Put the origin at 0 cm, choose right as positive, and place the robot at x = 20 cm.

  2. Move 30 cm right, so Δx = +30 cm. Predict x_new = 20 cm + 30 cm = 50 cm.

  3. Measure to check, then convert with 100 cm = 1 m: 50 cm = 0.50 m. Never add centimetres directly to metres.

Programmer analogy

A variable is like a named field in code, but a robot field must also say its physical unit and what zero means.

A robot starts at x = 40 cm and moves 15 cm left. What is its new x coordinate?

Left is negative, so Δx = -15 cm and x_new = 40 - 15 = 25 cm.

Always write a frame label, such as Ap^{A}\mathbf p, meaning “point p\mathbf p described in frame AA.” A simple right-handed axis set is

x^=(1,0,0),y^=(0,1,0),z^=(0,0,1).\hat{\mathbf x}=(1,0,0),\qquad \hat{\mathbf y}=(0,1,0),\qquad \hat{\mathbf z}=(0,0,1).

The hat means an axis arrow of length one. The right-hand rule fixes which way the three positive axes belong together.

Describe one coin in two aligned frames

A book-frame origin is at table x = 20 cm. Its +x axis is aligned with table +x. A coin is at book x = 5 cm.

  1. Name the frames: table T and book B.

  2. Record the book origin in T: x_TB = 20 cm.

  3. Record the coin locally in B: x_Bcoin = 5 cm.

  4. Check axis alignment: the book +x direction matches table +x, so the sign is +1.

  5. Add the aligned displacement: x_Tcoin = 20 cm + 5 cm = 25 cm.

  6. Verify with the ruler from the table origin.

Result

The same coin has coordinate 5 cm in the book frame and 25 cm in the table frame.

What this proves

Different coordinate values can correctly describe one physical point because origins and axes differ.

Physical examples

Where this appears in real life

Phone on a rotated book

A coin is 5 cm to the right of the book's centre; rotate the book 90° on the table without moving its centre.

Look for:

The coin's book-frame coordinates stay similar while its table-frame direction changes.

Camera mounted on a rover

A box appears one metre in front of a forward camera while the rover faces west in the map.

Look for:

Camera +x or +z forward, depending on convention, is not automatically map +x; the declared axis convention decides.

Hands-on exercise

Make the idea observable

Use a book, table, coin, paper axis arrows, ruler, and your right hand.

  1. Tape a 2D frame to the table and another to the book, with origins clearly marked.

  2. Place the coin at a measured book-frame coordinate and also measure it in the table frame.

  3. Rotate the book 90° around its marked origin without sliding that origin.

  4. Predict the coin's new table-frame direction before measuring it.

  5. Add a +z arrow to both frames and check the right-handed ordering.

  6. Deliberately reverse one y-axis and explain why the reflected frame is inconsistent.

Observe

Book-local coordinates follow the book, while table coordinates reflect the book's changed orientation.

Done when

Your note includes two origins, axis directions, units, coordinates before and after rotation, and a labelled reflected failure case.

Build today

Build a paper two-link robot and a browser notebook that shows its parts, allowed motion, coordinates, vectors, frames, and matrices.

Evidence to save

DONE when a 60–120 second uncut “Coordinate frames, axes, origins, and handedness” demo links to its command, logs or plots, result count, and honest failure note.

Common mistakes

Catch the wrong mental model

Wrong

Treating coordinates as properties that never change with the frame.

Better

The physical point is fixed, but its component values depend on the chosen origin and axes.

Wrong

Adding a camera-frame vector directly to a map-frame vector.

Better

Express both in one common frame first.

Wrong

Drawing a mirrored axis set and calling it a rotation.

Better

Check handedness; a reflection reverses the axis orientation and is not an ordinary rigid rotation.

Wrong

Using a correct transform with an old measurement.

Better

Match frame conversion to the measurement timestamp and reject stale data according to a declared bound.

Job connection

How this becomes employable evidence

Display and validate timestamped camera detections only after transforming them into the robot base or map frame used by planning and the operator interface.

Relevant target roles

  • Robotics Application / ROS 2 Integration Engineer
  • Robotics Software Engineer — ROS 2 / AMR
  • Robot HMI / Control & Monitoring Engineer

Chapter 01 interview drill

Interview questions: Coordinate frames, axes, origins, and handedness

Practise a 60–90 second answer: define the idea, connect it to a physical robot, state assumptions, frames, and units when relevant, then finish with the failure signal or evidence you would inspect.

Primary interview scenario

A detected obstacle looks correct in camera coordinates but appears behind the robot on the map. List the frame, axis, transform-direction, unit, and time checks you would perform.

Answer shape: clarify the situation → trace the physical and software path → test the most likely boundaries → name the evidence that would confirm the result.

Technical follow-up questions

Q1Can one coin have two correct coordinate pairs?
Model interview answer

Yes. Each pair can describe the same point relative to a different named origin and axis set.

Q2Why can you not add camera (1, 0) to map (1, 0) directly?
Model interview answer

The component directions and origins may represent different physical arrows; both must first be expressed in one frame.

Q3What information should accompany a position in a robot message?
Model interview answer

The coordinate values, unit, named reference frame, physical meaning, and measurement timestamp or freshness information.