T ThinkEdge Learn

Robotics and Autonomous Systems: The Four Sub-Areas

Robotics and autonomous systems is the field of machines that sense, decide and act in the physical world. For a student the useful question is not what the field contains, but which of its four parts an Arduino and a thirty dollar sensor can honestly reach.

robotics and autonomous systems robotics and autonomous systems journal autonomous systems research
Robotics · The four sub-areas, mapped to what a $30 sensor can do

The four sub-areas of the field

Perception turns the physical world into numbers: distance, position, image, sound. Planning turns those numbers and a goal into a sequence of actions, which is where path planning, behaviour trees and task decomposition live. Control turns a plan into motor commands, and it carries the most physics, because a controller has to cope with a machine that does not do exactly what it was told. SLAM, simultaneous localisation and mapping, cuts across all three: the robot builds a map of an unknown environment while working out where it is inside that map.

The papers sit on those boundaries too. IEEE Robotics and Automation Letters and the International Journal of Robotics Research cover the field broadly, the Journal of Field Robotics leans towards machines working outdoors, and the two big conferences, ICRA and IROS, are where most of this work first appears. A student reading a survey in one of those journals usually finds that the hard part is not the novel algorithm but the sensor and the evaluation.

What a thirty dollar sensor can and cannot do

An HC-SR04 ultrasonic sensor costs about four dollars and gives you a distance to anything that reflects sound. That is enough for obstacle avoidance, a parking sensor and a rough map of a corridor, because the geometry is simple and the environment is small. It is not enough for SLAM in most rooms, because the readings are noisy at angles, the beam is wide, and the sensor cannot tell a wall from a chair leg at the same distance.

The honest mapping for a classroom is this. Perception you can do properly with ultrasonics and IR reflectance. Planning you can do with state machines and simple grid logic. Control you can do with PID loops on motors and servos. SLAM you can observe and explain rather than implement. That is not a limitation to apologise for. It is the correct boundary for the hardware, and a student who knows where it lies learns more than one who follows a tutorial blindly.

Step by step

  1. Read one survey paper end to end

    Take a single review article from a robotics journal and list the system it describes: sensors, planner, controller, dataset. Seeing how a real paper is structured is worth more than a term of tutorials.

  2. Build the sensing layer

    Get an ultrasonic sensor and an IR reflectance sensor reading reliably and logged over serial. Everything else in the field depends on measurements you can trust.

  3. Implement a state machine

    Write the robot's behaviour as explicit states with transitions, such as searching, approaching and backing off. This is the simplest form of planning, and it scales further than nested if statements.

  4. Close a control loop

    Drive a servo or a motor to a commanded position and correct it from feedback. Add a proportional term, then integral and derivative, and watch each one change the response.

  5. Log everything

    Send every sensor reading and actuator command to a serial hub and record it on a PC. The report you write at the end of the project is a write up of a dataset you actually collected.

  6. Map the limits honestly

    List what the setup cannot do and why: beam width, minimum range, mechanical backlash, sampling rate. Naming a limitation precisely is what a researcher does, and it earns marks.

What you'll need

ItemQtyWhy
Arduino Mega 25601Four hardware serial ports, so the sensor bus and the logging link do not fight over one port.
HC-SR04 ultrasonic sensor2Two of them give a range and a bearing to a target, which is the minimum for naive obstacle avoidance.
IR reflectance sensor module2Fast, short range and immune to the sunlight patterns that upset the ultrasonic sensor.
L298N dual H-bridge module1Drives the two motors of the platform that carries the sensing layer around.
SG90 micro servo2Let a sensor pan and tilt, which is the cheapest way to widen the field of view.
9 V battery and barrel jack1Powers the motors separately, so motor noise does not corrupt the sensor readings.

Practise it before you build it

Every step above has a matching circuit in the ThinkEdge Learn simulator. The simulator runs a real Arduino sketch against a modified-nodal-analysis solver, so digitalWrite actually drives the LED on screen and the readings change when you change the wiring — no hardware, no burnt components.

Common questions

What is robotics and autonomous systems as a field?

It is the study of machines that perceive their environment, decide what to do and act on it without a human in the loop. The work sits in four areas: perception, planning, control and localisation, with SLAM cutting across them. Much of the published output appears in IEEE and IJRR journals and at the ICRA and IROS conferences.

What do I need to start doing research in robotics?

Less hardware than people expect and more maths than they expect. A decent laptop, an Arduino or a Raspberry Pi, two or three sensors and a well chosen dataset will support a genuine project. The skill that matters most is running experiments you can repeat and reporting the failures honestly, because that is what separates a research project from a demonstration.

Can an Arduino do SLAM?

Not real SLAM, but you can implement something that looks like it in miniature. A grid of cells marked visited, a robot that sweeps an ultrasonic sensor, and a simple rule for merging readings into occupied and free cells gives you occupancy grid mapping. It works in a corridor or a small room, and it falls apart as soon as the sensor stops seeing the surfaces it expects.

Related guides

How to Learn RoboticsHow to learn robotics without a degree: the order to learn electronics, code and control in, how long each stage takes, and the projects that prove each one landed. Robotics for BeginnersRobotics for beginners: five projects that take you from a blinking LED to a robot that steers itself, with the parts and the mistakes at each step. Autonomous Vehicles: What the Car Actually SensesAutonomous vehicles fuse lidar, radar, cameras and ultrasonics into one picture of the road. Here is what each sensor is good at, and how to build the loop. Humanoid Robots: Balance First, Walking SecondWhat makes a humanoid robot work: inverted-pendulum balance, joint torque budgets and centre of mass, plus the Arduino projects to get you there. How to Build a RobotHow to build a robot from scratch: one motor and a switch, then sensors, then a line-follower that steers itself. Every stage runs in a browser simulator first.

Try the full simulator with your class

103 guided circuit projects, 21 lessons and an in-browser Arduino simulator. The first 15 projects are free, no account needed. NCERT Science mapping on 47 projects, with Arabic and English copy in the regional builds.