Indoor robotics roadmap — from sensors to motion control

Contents — find the section you need

Background and reading guide

Background No robot-building experience required; linked articles introduce frames and estimation mathematics.

Learning guide · A system diagram and reading plan; no new hardware performance measurements.

An indoor wheeled robot becomes easier to understand when you follow one question: how does a sensor reading become a motor command? This guide connects the site's explanations and project records rather than asking you to memorize algorithm names.

A first reading goal

You do not need a robot or a ROS 2 installation to begin. Write four stages on paper: observe → estimate position → choose an action → issue a command. Name one input and one output for each stage. That is the starting goal of this page; equations and algorithm implementations can follow.

For the overall picture, continue with the diagram below. To work through communication, frames, estimation, mapping and planning in order, use the robotics learning guide. To begin with a small experiment in feedback response, use the control engineering guide to reach the PID explanation and browser experiment.

Follow the information flow

Observe, estimate, plan, control, and measure the resulting motion again

Commanded and actual velocity differ when wheels slip or loads change. Observation therefore repeats throughout motion. A stored map can also disagree with obstacles currently in front of the robot, so local observations must influence planning.

1: Understand what the sensors measure

Begin with LiDAR, IMUs and encoders. The useful milestone is being able to name what each sensor cannot determine alone. Wheel rotation does not directly measure slip against the floor, and integration of IMU measurements accumulates error.

2: Estimate the current pose

Read dead reckoning, LiDAR SLAM and sensor fusion. Separate a smooth short-term motion estimate from a pose corrected relative to a map. This makes the roles of odom and map easier to follow; their semantics are specified in ROS REP-105.

3: Represent free space and choose a route

Use occupancy grids for spatial representation and path planning for route selection. A point can pass through a gap that a wide robot cannot. As an exercise, explain why changing the robot footprint changes which corridors are traversable.

4: Turn the plan into motion

Study PID control and the browser experiment, then move from ROS 2 fundamentals to the minimal Nav2 procedure. The official navigation concepts provide another reference for component responsibilities.

Goal	First check	Continue with
Understand principles	Write down every block's inputs and outputs	Detailed explanations
Run a simulation	Set the initial pose and a goal on a map	Minimal Nav2 procedure
Improve hardware	Record repeated runs under matching conditions	SLAM evaluation and power diagnosis

Check the boundaries

Explain what stops and what reaches the motors when observations disappear. A boundary you cannot explain identifies the next topic to study. When configuring hardware, consult the Nav2 first-time setup guide.