Introduction to Robotics

Learning Objectives

By the end of this page, you should be able to:

• Define a robot in engineering terms and distinguish it from a simple automated machine
• Identify the four functional blocks every robot needs: sensing, actuation, control, and power
• Classify robots by structure (industrial, mobile, humanoid, aerial, underwater) and by autonomy level
• Explain how the sense-plan-act loop connects a robot's hardware to its behavior
• Recognize the main disciplines (mechanical, electrical, computer science) that robotics draws from and why none of them alone is sufficient

Quick Answer

A robot is a programmable machine that senses its environment, processes that information, and acts on the physical world to complete a task with some degree of autonomy. What separates a robot from an ordinary machine (like a washing machine or a conveyor belt) is this closed loop of sensing, deciding, and acting — a robot changes its behavior based on what it perceives, rather than just following a fixed mechanical sequence. Robotics matters because it sits at the intersection of mechanical design, electronics, and software: a poorly designed robot can be mechanically perfect and still fail because its control loop is unstable, or electronically flawless and still fail because its structure can't carry the required load. Understanding robotics means understanding how these three domains constrain each other.

What Actually Makes Something a Robot

Students often think "robot" just means "a machine that moves by itself." That's too loose — a garage door opener moves by itself, but nobody calls it a robot. The distinguishing feature is feedback-driven autonomy: a robot uses sensor data to make decisions in real time, and those decisions change what the actuators do next.

This gives every robot the same underlying loop, regardless of whether it's a six-axis welding arm or a Mars rover:

Sense → Plan → Act, repeated continuously.

• Sense: sensors (cameras, encoders, IMUs, force sensors, ultrasonic/LIDAR) convert physical quantities into electrical signals.
• Plan: a controller (microcontroller, embedded computer, or full PC) interprets those signals against a goal and computes what should happen next.
• Act: actuators (motors, hydraulics, pneumatics) convert the controller's commands into physical motion or force.

If you remove any one of these three, you no longer have a robot — you have a sensor network, a computer program, or a plain machine.

The Four Functional Building Blocks

Every robot, no matter how exotic, is built from four subsystems:

1. Mechanical structure — the links, joints, and frame that give the robot its physical form and determine its workspace and load capacity.
2. Actuators — the "muscles." DC motors, stepper motors, servo motors, and pneumatic/hydraulic cylinders convert energy into controlled motion.
3. Sensors — the "senses." They measure the robot's own state (joint angle, velocity) or the external world (distance, image, force).
4. Control system and power — the "brain and bloodstream." A microcontroller or computer runs the control algorithm, and a power supply (battery or mains) keeps everything running.

Why it matters: when you're troubleshooting a robot that isn't behaving correctly, this four-block model tells you where to look first. A robot that moves erratically might have a sensor giving noisy readings, a control loop with badly tuned gains, an actuator that's underpowered for the load, or a structural problem like a loose joint — the symptom looks the same, but the fix is completely different.

Common misunderstanding: students frequently assume more sensors or a "smarter" algorithm always improves a robot. In practice, adding sensors without proper calibration and fusion often makes performance worse, because noisy or conflicting inputs confuse the control system. A well-tuned robot with two good sensors usually beats a poorly integrated robot with six.

Classifying Robots

Robots are usually classified along two independent axes: structure and autonomy level. Mixing these up is a common source of confusion in exams.

By Structure

Type	Example	Defining Feature
Industrial/fixed-base	Welding arm on an assembly line	Bolted to a base, works in a defined workspace
Mobile ground robot	Warehouse AGV, robot vacuum	Moves through the environment on wheels/tracks
Legged/humanoid	Boston Dynamics Atlas, Honda ASIMO	Uses legs for locomotion, needs active balance control
Aerial	Quadcopter drone	Operates in 3D space, must manage flight dynamics
Underwater	ROV for pipeline inspection	Must handle buoyancy, pressure, and limited communication

By Autonomy Level

• Teleoperated: a human directly controls every movement (a bomb-disposal robot's arm).
• Semi-autonomous: the robot handles low-level control while a human sets high-level goals (a surgical robot that filters hand tremor but is still guided by the surgeon).
• Fully autonomous: the robot senses, decides, and acts without human intervention during task execution (a Mars rover driving between waypoints).

Real-world example: a Roomba is a mobile, fully autonomous robot for its cleaning task, but it has almost no manipulation capability. A da Vinci surgical system is a fixed-base, semi-autonomous robot with very high manipulation precision but essentially zero autonomous decision-making — the surgeon is always in the loop. Same broad category ("robot"), completely different design priorities.

Why Robotics Needs Multiple Disciplines

A robotics engineer needs a working knowledge of:

• Mechanical engineering: link design, material selection, load-bearing structures, gear trains.
• Electrical/electronics engineering: motor drivers, sensor interfacing, power management, signal conditioning.
• Computer science: control algorithms, path planning, perception, real-time software.

No single discipline is sufficient. A mechanically perfect arm with a bad control loop will oscillate or overshoot. A brilliant control algorithm running on an underpowered motor will never hit its target position. This is why robotics courses sit across several departments — the subject is inherently interdisciplinary.

Key Terms

Term	Definition
Robot	A programmable machine that senses, decides, and acts on its environment with some autonomy
Actuator	A device that converts energy (electrical, hydraulic, pneumatic) into controlled mechanical motion
Sensor	A device that measures a physical quantity and converts it into an electrical signal
Degree of freedom (DOF)	An independent way a robot's parts can move (rotate or translate)
End-effector	The tool or gripper at the end of a robot arm that performs the task
Teleoperation	Direct real-time human control of a robot's movements
Autonomy	The degree to which a robot can sense, decide, and act without human intervention
Sense-plan-act loop	The continuous cycle of measuring the environment, computing a decision, and executing an action

Common Mistakes

1. Misconception: "Any machine that moves on its own is a robot." Why it's wrong: this ignores the feedback loop. A washing machine runs a fixed timed sequence — it doesn't sense outcomes and adjust its behavior. Correct understanding: a robot must close the loop between sensing and acting; a fixed-sequence machine (open-loop automation) is not a robot even if it moves autonomously.

2. Misconception: "More autonomy always means a better robot." Why it's wrong: autonomy level is a design choice driven by the task, not a universal measure of quality. A surgical robot is deliberately kept semi-autonomous because full autonomy would be unsafe and legally untenable. Correct understanding: the right autonomy level depends on the cost of errors, the predictability of the environment, and regulatory/safety requirements — not on what's technologically possible.

3. Misconception: "Robotics is mostly a software/AI problem now." Why it's wrong: even the best control software cannot compensate for an underpowered actuator, a structure that flexes under load, or a sensor with too much noise for the task. Correct understanding: robotics is a systems problem — mechanical, electrical, and software constraints all interact, and a bottleneck in any one domain limits overall performance.

Comparison and Connections

Concept	Similar To	Key Difference
Robot	Automated machine (e.g., conveyor belt)	Robot closes a sense-decide-act feedback loop; automation typically runs a fixed sequence
Teleoperated robot	Remote-controlled toy	Teleoperated robots often provide force/sensory feedback to the operator; RC toys usually don't
Industrial robot arm	CNC machine	Both are precise and programmable, but a CNC machine typically follows a pre-computed toolpath without adapting to real-time sensor feedback about the workpiece
Mobile robot	Autonomous vehicle	An autonomous vehicle is a specific, safety-critical subclass of mobile robot operating in unstructured, shared human environments

Practice Questions

Recall

1. What are the four functional building blocks present in every robot? Answer guidance: mechanical structure, actuators, sensors, control system/power supply — explain the role of each in one sentence.
2. Name the three levels of robot autonomy and give one example of each. Answer guidance: teleoperated (bomb-disposal robot), semi-autonomous (surgical robot), fully autonomous (Mars rover).

Understanding

3. Explain why a washing machine is not considered a robot even though it operates without a human touching it during a wash cycle. Answer guidance: it runs an open-loop, pre-timed sequence and does not sense outcomes and adapt its behavior — no feedback loop.
4. Why does robotics require knowledge from mechanical, electrical, and computer science disciplines simultaneously, rather than being purely a coding problem? Answer guidance: mechanical structure limits load/speed, electronics limits sensing/actuation fidelity, software limits decision quality — a weakness in any one limits overall performance regardless of the others.

Application

5. You're asked to design a robot to sort recyclable bottles by color on a conveyor belt. Identify what sensor, actuator, and control approach you would use, and justify the autonomy level. Answer guidance: color camera or color sensor for sensing, pneumatic pusher or robotic arm as actuator, a simple classify-then-actuate control loop, and full autonomy since the task is repetitive and well-defined.
6. A hospital wants a robot to deliver medication trays between wards. Should it be teleoperated, semi-autonomous, or fully autonomous? Justify your choice considering safety and reliability. Answer guidance: fully autonomous is reasonable for navigation in a mapped, low-risk environment, but should include a manual override/semi-autonomous fallback for edge cases like blocked corridors or elevator failures — safety-critical decisions favor a human-in-the-loop option.

Analysis

7. Compare an industrial welding arm and a Mars rover in terms of structure, autonomy, and sensing needs. What does this tell you about how robot design depends on the environment? Answer guidance: welding arm is fixed-base, low autonomy (executes a fixed program), operates in a structured, predictable environment. Rover is mobile, high autonomy (must handle communication delay), operates in an unstructured, unpredictable environment — design choices follow directly from environmental constraints, not from a general notion of "better" robots.
8. A robot performs perfectly in simulation but fails on the physical hardware. List two plausible causes tied to the four functional building blocks and explain how you would isolate which one is at fault. Answer guidance: causes could include actuator saturation/backlash not modeled in simulation, sensor noise or calibration error, or structural flex under real loads. Isolate by testing each subsystem individually — e.g., command open-loop actuator moves and measure actual vs. expected position to separate actuator/structural issues from control software issues.

FAQ

Q1: Is a self-driving car a robot? Yes — it's a mobile robot specialized for road navigation. It has the same sense-plan-act loop (cameras/LIDAR sense, the software plans a path, motors and the steering/brake actuators act), just built for a very specific and safety-critical task.

Q2: Do I need to know how to code to study robotics? Programming is one of three core skill areas (along with mechanical design and electronics), but you don't need to be an expert in all three from day one. Most robotics programs build these skills in parallel.

Q3: What's the difference between robotics and automation? Automation broadly includes any system that reduces human intervention, including simple open-loop machines. Robotics specifically refers to systems with the sense-plan-act feedback loop and generally implies more flexibility and adaptability than fixed automation.

Q4: Why do industrial robots dominate manufacturing but humanoid robots are still rare? Industrial robots operate in structured, predictable environments where fixed-base arms with high precision are ideal. Humanoid robots must solve much harder problems — dynamic balance, unstructured environments, and general-purpose manipulation — which remain active research challenges.

Q5: Can a robot exist without sensors? Not in the modern engineering sense. A machine that runs a fixed motion sequence with no sensing (like an old-style pick-and-place cam mechanism) is classified as open-loop automation, not a robot, because it can't adapt when conditions change.

Quick Revision

• A robot = a programmable machine with a sense-plan-act feedback loop, not just "anything that moves on its own."
• Four functional blocks: mechanical structure, actuators, sensors, control system/power.
• Sensors convert physical quantities to electrical signals; actuators convert control commands into motion.
• Classification by structure: industrial/fixed-base, mobile, legged/humanoid, aerial, underwater.
• Classification by autonomy: teleoperated, semi-autonomous, fully autonomous — this is a design choice, not a quality ranking.
• Robotics is inherently interdisciplinary: mechanical, electrical, and computer science constraints all interact.
• A machine without feedback (fixed timed sequence) is automation, not robotics.
• Higher autonomy isn't automatically "better" — task risk and predictability determine the right autonomy level.
• Symptom-based debugging: erratic behavior can stem from sensors, control tuning, actuator limits, or mechanical issues — the four-block model tells you where to look.
• End-effector = the tool at the working end of the robot (gripper, welding torch, surgical instrument).
• Degrees of freedom describe the independent ways a robot can move — this becomes central in kinematics (next topic).

Related Topics

Prerequisites: Basic mechanics (forces, motion), basic electronics (voltage, current, simple circuits), introductory programming concepts.

Related Topics: Sensor Integration, Control Systems for Robotics, Robot Design and Construction.

Next Topics: Robot Kinematics — once you understand what a robot is and its building blocks, the next question is how to mathematically describe its motion.